Mitodru Niyogi
Gyan AI Research
mitodru@bharatgpts.com&Arnab Bhattacharya
Dept. of Computer Science & Engineering,
Indian Institute of Technology Kanpur,
India
& Gyan AI Research
arnabb@cse.iitk.ac.in
Abstract
In this paper, we present Paramanu-Ayn, a collection of language models based exclusively on legal case documents of India. In particular, the court case documents of the Supreme Court of India (SCI), and text of the Constitution of India and the Indian Penal Code are utilized. The novel Auto-Regressive (AR) decoder-only language model is pretrained from scratch at a context size of 8192 on a single A100-PCIE-40GB GPU with an efficient MFU (Model FLOPs Utilization) of 41.35. Our model size is only 97 million parameters. We also trained a tokenizer specialized for legal corpora based on BPE. We evaluated our pretrained legal model on perplexity metric, and did zero-shot evaluation on two tasks: case judgement prediction with explanation and abstractive legal case summarization. Our pretrained legal model outperformed LlaMa-2 7B and Gemini-Pro on zero-shot test accuracy for the legal case judgement prediction with explanation task by close to 2% points, despite being smaller in size by close to 72 times compared to these 7B LLMs. We also reported automatic lexical metrics of zero-shot abstractive summarization generation of our Paramanu-Ayn model across varying summary lengths and compared with decoder-only LLMs at fixed summary length generation (5000 tokens) using ROUGE-1, ROUGE-L, BLEU, METEOR, and BERTScore metrics. Paramanu-Ayn outperformed the models in both BLEU and METEOR metrics by more than 10% percentage points, and also on BERTscore by close to 4% points.Furthermore, we evaluated our model on popular LLM commonsense and mathematical benchmarks for a holistic evaluation despite the model being pretrained exclusively on legal court cases documents. Paramanu-Ayn scored impressively on all the benchmarks. Paramanu-Ayn outperformed Llama-1, Llama-2, and Falcon LLMs on mathematical benchmarks (AGIEVAL-AQuA-RAT, AGIEVAL-SAT-Math) at zero-shot setting.We also instruction-tuned our pretrained model on a set of 10,763 instructions covering various legal tasks such as legal reasoning, judgement explanation, legal clause generation, legal drafting, legal contract drafting, case summarization, constitutional question-answering, etc. We also evaluated the responses of prompts for instruction-tuned models for both legal instructions related to Supreme Court of India cases and on drafting legal clauses and legal contracts by GPT-3.5-Turbo on clarity, relevance, completeness, and legal reasoning metrics. Paramanu-Ayn-instruct model scored more than 8 (on a scale of 10) on all the metrics, namely, clarity, relevance, completeness, and legal reasoning.Our model can be run on CPU and achieved 42.46 tokens/sec CPU inference speed. We found that our models, despite being pretrained exclusively only on court case documents, were able to learn the domain knowledge required for drafting various legal contracts and legal clauses, and generalize to draft legal contracts and legal clauses with limited instruction tuning. Hence, we conclude that for a strong domain-specialized generative language model (such as legal), small domain specialized pretraining from scratch is more cost effective. Our 97 million parameters model was pretrained from scratch for only 185 A100 training hours but is still equally competitive or even better than adapting LLMs for legal domain tasks. We believe that this work is the first attempt to make an exclusive generative legal language model from scratch for Indian Supreme Court jurisdiction or in legal NLP overall, and offer an environmental friendly cost efficient approach for legal domain adaptation.
1 Introduction
Legal text is characterized by its unique syntax and specialized vocabulary. This poses a distinct linguistic challenge for language models (Chalkidis etal., 2020; Niklaus and Giofre, 2023). Pretrained Large Language Models (LLMs) such as LLaMa (Touvron etal., 2023), Longformer (Beltagy etal., 2020), etc. are often trained on domain specific data for domain adaptation (Yao etal., 2021; Gururangan etal., 2020; Colombo etal., 2024). We propose a different approach. Instead of following the domain adaptation method of LLMs for legal language modeling for legal documents,we focused to pretrain from scratch a generative legal language model for case documents of Supreme Court of India.This avoids high inference latency, high cost of training, and non-specialised tokenizer and the misalignment of domain specialised tokenizers and embeddings with the existing embeddings of large language models (LLMs) via continual pretraining with vocabulary expansion of the existing LLMs tokenizers.Instead of developing humongous models that are believed to be good for all domains and all tasks, we believe in distributed specialized domain-expert machines. Hence, we want to develop domain-specific smaller models from scratch for resource-constrained computing setting.
Our model is based on the Transformer Decoder architecture Vaswani etal. (2017). We have trained a collection of an auto-regressive generative legal language models from scratch at a context size of8192on a single NVidia A100-PCIE-40GB GPU.Our work is an attempt to make dedicated domain specialized models from scratch rather than performing continual pretraining of existing LLMs for domain adaptation. Our models are much smaller in size and have just 97 million parameters. Hence, our models are very fast in inference without requiring any quantization of weights. Our legal models can be run on CPUs or even in smartphones without the need of GPUs.
Our main contributions in this work are as follows:
- •
We have created a pretraining corpus with the Constitution of India, Indian Penal Code and case documents from Supreme Court of India till December, 2023.
- •
We developed the first ever dedicated collection of generative legal AR models pretrained from scratch for Indian Supreme Court at context size of 4096 and 8192 respectively on a single GPU.
- •
We evaluated our legal pretrained models on validation perplexity, and on model FLOPs Utilization (MFU) metric for pretraining. Table2 shows the metrics.
- •
We performed zero-shot evaluation of our legal model on the case judgement prediction with explanation task. Table3 shows that our model outperformed LLMs such as LLaMa-2 7B and Gemini-Pro on zero-shot evaluation.
- •
We performed automatic lexical evaluation of zero-shot summary generation of our legal model and decoder-only LLMs on the case summarization task.Table4 shows the results of the evaluation of Paramanu-Ayn across abstractive summary generation across varying lengths of token generation.Table5 compares the automatic lexical metrics of zero-shot evaluation of LLMs versus Paramanu-Ayn on zero-shot case abstractive summarization task.
- •
We also evaluated Paramanu-Ayn on zero-shot evaluation on common LLM benchmarks across commonsense reasoning, logical, analytical, and mathematical reasoning.Table6 shows the benchmark evaluation of Paramanu-Ayn on these tasks. On the mathematical benchmark, AGIEVAL-AQuA-RAT, and AGIEVAL-SAT-Math, Paramanu-Ayn 97M outperformed various LLMs like Falcon (7B,40B), Llama-1 (7B, 13B, 33B), Llama-2 (13B, 34B) respectively. For abstract reasoning on AGIEVAL-LSAT-AR, Paramanu-Ayn 97M also outperformed Falcon and Llama LLMs.
- •
We further instruction-tuned our Paramanu-Ayn legal pretrained model on 10,763 legal instructions and demonstrated various legal task capabilities.
- •
We evaluated the responses of Paramanu-Ayn-instruct by GPT-3.5-Turbo on clarity, relevance, completeness, and legal reasoning metrics.Table8 shows the results.
2 Related Work
Large Language Models (LLMs) have predominantly trained on variety of data ranging from web corpus, books to scientific articles, and also on legal data but majorly exposed to Western legal data. However, legal data typically represent a small fraction of the entire pretraining dataset. A common approach for domain adaptation of LLMs in legal NLP have traditionally been to do additional training to enhance the performance of LLMs for legal tasks. Notable ones are Chinese legal LlawyerLlama (Huang etal., 2023), LLaMa trained on a large-scale legal dataset, and ChatLaw (Cui etal., 2023), a legal LLM with Integrated External Knowledge Bases. Both of the said models rely on retrievers to extract relevant external knowledge and generate the required answers using LLM.
Talking about legal evaluation, of law-specific LLMs needs benchmarks that focus on legal tasks. Since legal laws and processes vary by country and country specific, developing a single global legal LLM evaluation benchmark may not be desirable in our opinion.
Legalbench Guha etal. (2023) is a collaboratively constructed legal reasoning benchmark comprising 162 tasks in the USA law context. LawBench (Fei etal., 2023) is a similar benchmark in the Chinese context. However, such legal benchmarks in the Indian context do not exist to the best of our knowledge.
3 Background
3.1 Language Modeling
This objective of the language modeling can be formally described as maximizing the probability of a sequence of tokens
where is the probability of token given the sequence of previous tokens .
The performance of a language model is generally evaluated using the total cross-entropy loss, i.e, the negative log-likelihood of the observed data under the model under consideration, which for a givendataset is defined as:
Lower the loss better is the model; however, just computing the loss may not be intuitive. Therefore, Perplexity is a metric to evaluate the performance of a given language model which is the exponent of the average loss.
3.2 Rotary Position Embedding (RoPE)
Transformer-based models rely on positional embeddings toencode position and relative location information of words in a text.Rotary Position Embedding (RoPE) is a position encoding technique proposed by (Black etal., 2022).Instead of adding positional embeddings or relative positional embeddings to token embeddings, RoPE rotates the token embedding by a fixed factor () in the higher-dimensional space to encode relative positional embeddings. In other words, RoPE encodes the absolute positions with a rotation matrix and meanwhile incorporates the explicit relative position dependency in self-attention formulation. The intuition behind RoPE is that we can represent the token embeddings as complex numbers and their positions as pure rotations that we apply to them. By exploiting the nature of rotations, the dot product used in self-attention will have the property for preserving relative positional information while discarding absolute position.
3.3 Root Mean Square Normalization (RMSNorm)
To improve the training stability, some LLMs (Chinchilla (Hoffmann etal., 2022), LLaMa (Touvron etal., 2023)) have normalized the input of each transformer sub-layer, instead of normalizing the output using RMSNorm normalizing function as introduced by (Zhang and Sennrich, 2019).RMSNorm normalizes the activations based on their root mean square (RMS) value instead of normalizing the inputs based on their mean and variance.RMSNorm accelerates the training and inference with similar performance in these large models. It is reported that replacing LayerNorm (Ba etal., 2016) with RMSNorm can achieve comparable performance and improve training and inference time by 7-64%.Narang etal. (2021) showed that RMSNorm improves the pre-training speed by 5% compared with the LayerNorm baseline.
3.4 Model FLOPs Utilization (MFU)
Model FLOPs Utilization (MFU) (Chowdhery etal., 2024) estimate is the ratio of the observed throughput (tokens-per-second) relative to the theoretical maximum throughput of a system at peak FLOPs. Model flops utilization (MFU) estimate the number of flops (floating point operations) done per iteration. It quantifies how efficiently the GPUs are utilized in model training.
3.5 Maximal Update Parameterization
As the size of large language models (LLMs) and the scale of the dataset used in pretraining are expensively large, it is not feasible to perform hyperparameter tuning in LLMs. Yang etal. (2021) used a technique called maximal update parameterization () to transfer the hyperparameters learnt from tuning of a small model to a larger model and found that the optimal hyperparameter values become stable across neural network sizes when the models have been parameterized using ().
4 Data
4.1 Pretraining Data
The pretraining data is an extended version of ILDC (Malik etal., 2021) dataset which initially contained 34,816 Supreme Court of India cases. We extracted all the publicly available case proceedings from May 2020 to December 2023 (3,046 case documents) and augmented the dataset.Apart from Supreme Court cases, we also have added the Constitution of India (in English) and the Indian Penal Code (in English), which we scraped using public sources. Table1 displays the statistics for our clean preprocessed pretraining corpus. As the number of tokens vary due to the variation of tokenizer, we reported the total number of (whitespace-separated) words instead.
4.2 Data Cleaning and Preprocessing
Case proceedings are unstructured documents and have different formats and sizes. In addition, there are typos and spelling mistakes (since these aretyped during the court hearing), thereby making it challenging to preprocess. We used regular expressionsto remove the noisy text (unreadable characters, special characters, unnecessary whitespaces, etc.) and meta-information (e.g., initial portions of the document containing casenumber, judge name, dates, and other meta information) from the proceedings.In Supreme Court case proceedings, the decisions arewritten towards the end of the document. These endsection(s) directly stating the decision have beenkept instead of removing from the documents like the ILDC dataset.
For both the Constitution of India and the Indian Penal Code, we used the tesseracttool to extract the text from the pdf files and also removed the first few pages of the pdf to remove title page, acknowledgments, table of contents, and other irrelevant information.
Pre-training Corpus Timeline #Words 38,222 SCI Cases 1947 – 2023 142,374,414 Constitution of India 2024 139,956 Penal Code of India 2024 77,633 Total 1947 – 2024 142,592,003
4.3 Instruction Tuning Data
The instruction dataset is a cumulative corpus of various existing publicly available datasets (NebulaSense, 2023; Ayansk11, 2023).This results in a total of 12,270 instructions; after removing duplicates, it resulted in 10,763 instructions covering various legal tasks such as legal memo drafting, legal petition drafting, legal contract generation such as employee contract drafting, NDA drafting, legal clause generation, legal argument challenging, legal advice based on an outcome of a Supreme Court case, legal case summarization and key findings explanation, legal precedents identification, legal question answering, etc.
5 Tokenization
We trained Byte-Pair encoding (BPE) (Sennrich etal., 2016) algorithm using Sentencepiece (Kudo and Richardson, 2018) module on the pretraining data from scratch to develop legal specialised tokenizer that aims to learn the intricacies of legal terminology. During pre-tokenization, NFC normalizationwas performed on the processed data; digits are split into individual tokens and unknown UTF-8 characters were reduced to byte granularity for improving the arithmetic learning ability of the pretrained model.In NFC, characters are normalized by composing them into their canonical form. This means that composed characters (like “é”) are represented as a single code point rather than as a combination of base characters and diacritical marks (like ‘e’ followed by an accent mark). This helps in ensuring consistency and proper comparison of text, as different sequences of characters that visually represent the same text can be treated as identical.We treat the data as a sequence of bytes rather than Unicode characters, and we include each of the 256 bytes as tokens too. We removed the overlapping tokens from the trained tokenizer by removing the duplicated tokens in the tokenizer to develop our tokenizer legal specialised, compact, optimized, and effective. The size of the final tokenizer vocabulary is 15,575.
6 Model Architecture
The model architecture of Paramanu-Ayn uses a scaled version of RoPE embedding (Su etal., 2024) with =10,000.We scaled the RoPE embedding through a shrinking factor by dividing the target context length by on single GPU, keeping all other hyperparameters fixed such as batch size, vocabulary size, etc. This allows every to be divided by the shrinking ratio in the RoPE embedding methodology. For instance, if the for a given physical memory hardware is 256, then we apply shrinking factor of 16 for target context size of 4096 on Single A100 40G chip during pretraining. Then, a token with = 4000 becomes 4000/16 = 250, and the neighbouring token 4001 becomes 4001/16 = 250.06, to be within 0 to 256. This is how we can capture higher context size during pretraining on limited physical memory required to pretrain model at higher context size outside the . This modification allows us to pretrain models from scratch at much higher context size than the physical memory required for pretraining. Hence, with limited physical memory and limited GPUs, we can pretrain generative language models from scratch at much higher desired context size. Applying this technique allowed us to pretrain our legal model at a context size of 8192 on a single NVIDIA A100-PCIE-40GB GPU. The shrinking factor was set to 32 to achieve a target context size of 8192.
The model architecture of Paramanu-Ayn uses RMSNorm (Zhang and Sennrich, 2019) as a pre-normalizaion layer with norm_epsilon = 1e-5, uses SwiGLU (Shazeer, 2020) an activation function for non-linearity by replacing the standard ReLU activation function. The dimension of Paramanu-Ayn is 768 with 12 layers with a feedforward layer hidden dimension of 2048. Following (Su etal., 2024), we remove all biases from dense layers toimprove the training stability of Paramanu-Ayn.The model Paramanu-Ayn uses grouped-query attention (GQA) (Ainslie etal., 2023) to reduce inference-time memory footprint and accelerate inference at run time. GQA divides query heads into groups, each of which shares a single key head and value head. GQA-G refers to grouped-query with groups. In GQA, we divide the total self-attention heads into groups, where key and value heads are shared within each group; GQA-1, with a single group and therefore single key and value head, is equivalent to multi-query attention (MQA) (Shazeer, 2019), while GQA-H, with groups equal to number of heads, is equivalent to multi-head attention (MHA). We also used weight tying (Press and Wolf, 2017) to improve the performance of language models by sharing the weights of the embedding and softmax layers.
7 Training
For pretraining Paramanu-Ayn (8192 context size), we performed 95%-5% data split for pretraining, as we wanted to use most of the dataset for pretraining. We reported the validation perplexity of our pre-trained model. Figure1 shows the training perplexity curve against the number of tokens.

We performed hyperparameter tuning on 15M models to find the optimal vocabulary, learning rate, and warm-up ratio. We used a batch size of 8, gradient accumulation steps of 8, and the maximum sequence length set to 8192, i.e., 524,288 tokens per iteration. We used the concept of transfer, and transferred the learned hyperparameters to our bigger model for 97M Paramanu-Ayn from 15M model. Following (Hoffmann etal., 2022), we set decay steps to and the minimum is set to 0.1. The schedule starts with a linear warm-up from 0 to the maximum at 1000 steps, followed by a cosine decay to the minimum until the = 100,000 of training. We used the following equation for
where is the current training step.We set the maximum learning rate () to 0.003 (max), weight decay to 0.1. Paramanu-Ayn is pretrained on around a total of 26 billion tokens. Fig1 shows the plot of training perplexity with respect to every billion tokens.We trained with fused AdamW optimizer with = 0.9, = 0.95, a gradient clipping of 1.0, and weight decay of 0.1. To further speedup training, we used BF16 mixed precision training. Paramanu-Ayn was trained for 185 A100 hours. For our experiments and modeling, we implemented our code using Pytorch 2.0, in-house optmized CUDA kernels and used torch.compile feature for every model.
For instruction-tuning, we split the dataset into 90%-10% training and testing sets. Since our models are smaller in size, we performed supervised full fine-tuning of Paramanu-Ayn for 3 epochs on the accumulated 10,763 instructions and trained in different setting of experiments using cosine, constant, and linear learning rate scheduler with () set to e, gradient clipping of 1.0, warmup ratio of 0.05 and no weight decay.We found that cosine learning rate scheduler results in the lowest validation loss for our models.
8 Carbon Footprint
To measure carbon footprint for our pretraining, we follow Touvron etal. (2023):
The power consumption can be calculated as
where PUE is Power Usage Effectiveness.
We observed during pretraining that our single A100 40G consumes 250 Watt consistently. Therefore, our pretraining process consumed 50.875 kW, resulting in only 0.0196 tCO2eq respectively. This makes our model way more environmental-friendly as compared to pretraining/continual pretraining of LLMs.
9 Evaluation
We evaluated Paramanu-Ayn and LLMs including (a)LlaMa-3 8B, (b)Mixtral 7B, (c)Gemini Pro on (1)Legal Judgement Case Prediction task (Nigam etal., 2024) using zero-shot evaluation which is exclusively for Supreme Court of India cases, (2)Supreme Court of India abstractive summarization (In-Abs) (Shukla etal., 2022) test dataset of 100 case documents using zero-shot evaluation on summary generation of 5000 tokens. We further studied the long summary generation capability of Paramanu-Ayn by generating summaries of varying lengths of 1024, 2048, 4096, 5000, 6000, and 8192 tokens and compared the lexical evaluation metrics like ROUGE-1, ROUGE-L, METEOR, BLEU, and semantic evaluation metrics like BERTScore.ROUGE family of metrics measure the textual overlap (uni-grams) between the model-generated summaries and the reference summaries.METEOR calculates the harmonic mean ofunigram precision and recall. For summary evaluation, METEOR is used to calculate the unigram overlap between a model-generated summary and the gold standard summary. BERTScore uses BERT to compute the similarity scores between the token level embeddings of the model-generated and reference summaries.We also evaluated Paramanu-Ayn on zero-shot evaluation on MMLU-legal tasks (Professional Law, Jurisprudence, International Law) and analytical reasoningsuch as AGIEVAL-LSAT-LR (Law school admission tests) which measure the reasoning and analytical skills of prospective law students. These tests include sections on logicalreasoning, reading comprehension, and analytical reasoning.Furthermore, we went ahead to explore Paramanu-Ayn capabilities on various commonsense reasoning benchmarks such as ARC-easy, ARC-challenge, SciQ, PIQA, WSC, COPA, mathematical reasoning such as AGIEVAL-AQuA-RAT (GRE, GMAT multiple-choice math questions), AGIEVAL-SAT-Math (SAT multiple-choice math questions) (Zhong etal., 2024), AGIEVAL-MATH (Hendrycks etal., 2021b) (formal logical reasoning such as MMLU-Formal-Logic, causal reasoning task like COPA (selecting a conclusion causally connected to a premise), language coreference resolution task such as WSC (Levesque etal., 2012), classification task such as WIC (Pilehvar and Camacho-Collados, 2019) and natural language inference benchmark such as QNLI (Wang etal., 2018). We also reported the scores on zero-shot evaluation on MMLU-STEM, MMLU-Econometrics, and across all MMLU tasks.
For Paramanu-Ayn-instruct, we evaluated the responses of the test instructions-tuning dataset using GPT-3.5-Turbo on clarity, relevance, completeness, and legal reasoning in a scale of 10.
10 Results and Analyses
Model Perplexity MFU Inference Speed Paramanu-Ayn 1.4582 41.3476 42.4575 tokens/sec
10.1 Perplexity and Inference Speed
Table 2 reports the validation perplexity of our pretrained legal models, the model FLOP utilization (MFU) metric for model pretraining, and CPU inference speed in tokens/sec for Paramanu-Ayn pretrained model.
Our validation perplexity is 1.4582 which is a sign that the legal pretrained model has been trained well. Moreover, we reach MFU of 41.3476 at bf16 mixed precision training using our in-house implementation of optimized CUDA kernels for distributed training framework. This also indicates to our training efficiency.
10.2 Legal Case Judgement Prediction with Explanation
Model | Macro-F1 | Accuracy (%) |
---|---|---|
Gemini Pro | 0.4908 | 50.81 |
LLaMa-2 7B | 0.3772 | 50.25 |
Paramanu-Ayn | 0.5037 | 52.00 |
Length ROUGE-1 ROUGE-L METEOR BERTScore BLEU 1024 0.2458 0.1145 0.1548 0.4795 0.1000 2048 0.2472 0.1121 0.1974 0.5287 0.2260 4096 0.2055 0.0943 0.2080 0.5142 0.1767 5000 0.2387 0.1072 0.2400 0.4848 0.1529 6000 0.1835 0.0854 0.2060 0.5048 0.1453 8192 0.1635 0.0768 0.1964 0.4795 0.2035
Model ROUGE-1 ROUGE-L BLEU METEOR BERTScore Gemini Pro 0.2484 0.1531 0.0171 0.1141 0.5759 Mixtral 7B 0.2998 0.2910 0.3022 0.3723 0.4469 Llama-3 8B 0.2937 0.1821 0.0249 0.1338 0.6204 Paramanu-Ayn 0.2387 0.1072 0.1529 0.2400 0.4848
Task Dataset Accuracy (Type) Accuracy (Value) ARC-challenge Normalized 26.45 HellaSwag Normalized 27.18 MMLU Raw 25.30 MMLU-STEM Raw 24.10 MMLU-Formal-Logic Raw 33.33 MMLU-Jurisprudence Raw 24.07 MMLU-Professional Law Raw 23.66 MMLU-International-law Raw 19.83 MMLU-Econometrics Raw 32.46 TruthfulQA-mc2 Raw 51.81 WinoGrande Raw 50.28 PIQA Raw 52.18 SciQ Normalized 40.60 COPA Raw 54.00 WIC Raw 48.75 WSC Raw 62.50 QNLI Raw 50.92 WNLI Raw 59.15 MathQA Raw 20.10 AGIEVAL-LSAT-AR Normalized 22.61 AGIEVAL-SAT-Math Raw 27.27 AGIEVAL-MATH Raw 2.90 AGIEVAL-AQuA-RAT Raw 24.02 LogiQA Normalized 23.33

Model Size AQuA-RAT SAT-Math MATH LSAT-AR LSAT-LR MPT 7B 27.6 23.6 3.0 18.7 21.2 MPT 30B 28.0 23.9 23.9 35.1 Falcon 7B 21.7 26.4 2.3 16.1 17.3 Falcon 40B 18.5 32.7 5.5 19.6 40.2 Paramanu-Ayn (ours) 97M 24.02 27.27 2.90 22.61 20.20 Llama-1 7B 18.9 22.3 2.9 26.1 19.2 Llama-1 13B 20.1 29.5 3.9 22.2 31.6 Llama-1 33B 18.9 35.0 7.1 18.7 48.0 Llama-1 65B 23.6 41.8 10.6 23.9 56.7 Llama-2 7B 23.2 28.2 2.5 23.9 22.4 LLama-2 13B 21.7 27.3 3.9 23.0 41.0 Llama-2 34B 19.3 32.7 6.24 21.3 47.5 Llama-2 70B 23.2 41.8 13.5 25.7 70.2
We used the following zero-shot inference prompt for case judgement prediction with explanation:
“Analyze the case proceeding and predict whether the appeal/petition will be accepted (1) or rejected (0).
Subsequently provide an explanation behind this prediction with important textual evidence from the case.
### Input: case_proceeding:
### Response:”
Table3 compares Paramanu-Ayn with LLMs on the case prediction with explanation task. It outperformed both LLMs significantly in both the test accuracy and macro-F1 scores despite being much smaller in size by 80 times compared to 7B LLMs and was trained for only 185 A100 hours. We also would like to point out that for domain adaptation, if a tiny generative language model is pretrained from scratch sufficiently enough, can become very competitive or even better given our findings is much more cost efficient, environmentally friendly and faster in inference by multiple order of magnitude without any optimization due to the reduced size in model parameters compared to LLMs, and reduced carbon footprint than continual pretraining of LLMs or training adapters for domain adaptation and merging with LLMs which increases the latency and inference cost drastically even one might argue on the training cost of adapters for domain adaptation.
10.3 Legal Case Abstractive Summarization
We used the following zero-shot inference prompt for case summarization:
“You are a legal assistant and your job is to summarize the underneath case proceeding given in a most concise manner while being safe.
Your summary must have the same meaning and not include false information. Make sure you do not use any external knowledge other than what is provided to you.
Your final output must only be the summarized text.
### Case:
### Summary:”
Table5 shows the comparison of automatic summarization evaluation metrics of zero shot summary generation of Paramanu-Ayn versus LLMs at a summary generation length of 5000 tokens.It can be seen that Paramanu-Ayn outperforms most of the models for most of the metrics.Despite being trained only on case documents, the model has relatively performed well and even better than much larger LLMs which are trained on all kinds of variety data. This means that preprocessed legal case documents are high quality pretraining data that results in models acquiring different capabilities.
10.3.1 Impact of Length of Abstractive Summary Generation
Table4 shows the automatic summarization evaluation of zero shot summary generation of Paramanu-Ayn 97M across varying length of abstractive summary generation. We observe that Paramanu-Ayn scored the highest BERTScore of 0.5287 at summary generation length of 2048 tokens. ROUGE-1 and ROUGE-L scores are highest for the 2048-token summary, which reflects the best match in terms of unigram overlap and sequence length with the reference summary. The METEOR and BLEU scores are also strong, suggesting a better balance between precision, recall, and n-gram overlap compared to shorter summaries.
While longer summaries tend to perform well on some metrics (e.g., BLEU and METEOR), they show a decline in ROUGE-1 and ROUGE-L scores. This suggests that extremely long summaries may lose coherence or relevance compared to the reference summary, despite capturing more information. The METEOR score improves with longer summaries up to 5000 tokens, indicating that the model’s precision and recall enhance with the length. However, very long summaries (8192 tokens) show a slight decline, suggesting a possible trade-off between completeness and accuracy. The BLEU score increases with summary length, peaking at 2048 tokens and remaining relatively high for longer summaries. This reflects better n-gram overlap but also suggests that excessively long summaries might introduce redundancy or irrelevant content.
10.4 Common-Sense Reasoning and Mathematics
Table6 reports the zero-shot evaluation of Paramanu-Ayn across various commonsense, causal, complex reasoning tasks of natural language understanding.Paramanu-Ayn performs better than the random baseline of 25% on MMLU, i.e, a performance higher than GPT-3 models of size 13 billion parameters on zero-shot as reported in (Hendrycks etal., 2021a).Paramanu-Ayn 97M scored an impressive score of 26.45 on ARC-challenge benchmark, 25.3 on MMLU, 51.81 on TruthfulQA-mc2, 24.02 on AGIEVAL-AQuA-RAT, 27.27 on AGIEVAL-SAT-Math via zero-shot evaluation despite being exclusively pretrained on only legal case documents and having only 97 million parameters. On the mathematical benchmark, AGIEVAL-AQuA-RAT benchmark, Paramanu-Ayn 97M outperformed various models: Falcon 7B by 2.32%, Falcon 40B by 5.12% and Llama-1 7B by 5.32%, Llama-1 13B by 3.92%, Llama-1 33B by 5.32%, Llama-2 13B by 2.32%, Llama-2 34B by 4.72% points respectively. On AGIEVAL-SAT-Math, Paramanu-Ayn outperformed Falcon 7B and Llama-1 7B by 4.97%. For abstract reasoning on AGIEVAL-LSAT-AR, Paramanu-Ayn 97M outperformed Falcon 7B by 6.51%, Falcon 40B by 3.01%, Llama-1 33B by 3.91%, and Llama-2 34B by 1.31% points. The scores for Llama-1, Llama-2, and Falcon models are taken from the Llama-2 paper (Touvron etal., 2023), which used 3-5 shot evaluations for AGIEVAL. In contrast, our model was evaluated using at zero-shot setting.
TruthfulQA assesses the model’s ability to provide accurate answers, emphasizing its grasp of factual information and its capacity to avoid falsehoods (Lin etal., 2022). MMLU measures the model’s broad knowledge across various subjects including humanities, science, technology, engineering, and management (Hendrycks etal., 2021a). ARC-Challenge tests the model’s ability in complex reasoning with scientific questions (Clark etal., 2018). PIQA assesses the model’s knowledge of everyday physical processes, requiring understanding of physical commonsense (Bisk etal., 2019). COPA (The Choice of Plausible Alternatives) is a causal reasoning task for English. Given a premise and two alternatives, the task is to select the alternative that more plausibly is either the cause or effect of thepremise (Gordon etal., 2012).
Paramanu-Ayn performed relatively well, achieved 62.5%on on the Winograd Schema Challenge (WSC), indicating a strong grasp of contextual and commonsense reasoning tasks. It scored 51.81% on Truthfulqa-MC2 showing reasonable accuracy in evaluating the truthfulness of statements, demonstrating effective handling of complex language understanding. It scored 52.18% on PIQA suggesting good problem-solving skills in commonsense reasoning, particularly for questions requiring everyday knowledge. Paramanu-Ayn demonstrated good performance on causal reasoning task in English; it scored 54% in COPA in zero-shot setting.
10.5 GPT-3.5-Turbo Evaluation
Table8 shows the GPT-3.5-Turbo (ChatGPT) evaluation of Paramanu-Ayn-instruct model’s responses to legal instructions related to various drafting clauses, modifications, and legal contracts across various industries on different metrics such as clarity, relevance, completeness, and legal reasoning. For GPT-3.5-Turbo evaluation, we queried ChatGPT with the respective instruction, input and our model responses from the test set to evaluate the model responses on clarity, relevance, completeness, and legal reasoning in a scale of 10. Despite Paramanu-Ayn not being pretrained on legal books and legal contracts of various types, but only exclusively on Supreme Court Case documents and with limited instruction-tuning on legal contracts, legal clauses and modifications instructions, scores (out of 10) of 6.77 on clarity, 7.75 on relevance, 7.50 on completeness, and 7.75 on legal reasoning could be achieved. In our humble opinion, this shows the general learning ability of our models from limited instruction tuning. Table 8 evaluates the same but for only various legal instructions related to Supreme Court Cases in India.It shows significant improvement, with all scores of 8 and above.This improvement in the scores leverages the importance of pretraining with relevant corpus.
Task Clarity Relevance Completeness Legal Reasoning Legal clauses and contracts 6.75 7.75 7.50 7.75 Supreme Court of India (SCI) 8.00 8.88 8.22 8.89
10.6 Case Studies
We performed 18 kinds of legal tasks for instruction tuning.They are listed with examples in Supplementary Material.
11 Conclusions
In this paper, we proposed an alternative approach for domain generalization, instead of doing continual pre-traning of LLMs or training adapters with LLMs on respective domain such as legal for domain adaptation or generalization, we pretrained a tiny generative language model from scratch for only 185 A100 hours of training. We presented an exclusive Indian legal auto-regressive decoder based generative language legal model, Paramanu-Ayn, pretrained from scratch exclusively only on Supreme Court cases in India, Constitution of India, and Indian Penal Code for a context size of 8192. We evaluated Paramanu-Ayn on zero-shot setting for two main legal tasks: case judgement prediction with explanation, and abstractive case summarization. Our pretrained legal model outperformed much larger LLMs trained on legal data on most of the metrics.We found out that pretraining a competitive domain specific tiny language model is more cost efficient and faster in inference by multiple order of magnitude without any optimization due to the reduced size in model parameters. If a tiny language model trained sufficiently enough, then pretraining such models for domain adaptation is much more cost efficient with low latency in inference, environmental friendly due to very less carbon footprint generation and yet very competitive to LLMs and much cost efficient, faster than continual pretraining or merging adapters to LLM which in turn increases the model parameters and increases the inference cost and latency, is indeed a new exciting research direction for domain adaptation.Furthermore, we evaluated our model on zero-shot setting across popular LLM commonsense and quantitative benchmarksfor a holistic evaluation despite the model being pretrained exclusively on legal court cases documents.We also instruction-tuned our legal model, i.e, Paramanu-Ayn-instruct on 10,763 legal instructions curated from publicly available datasets, and demonstrated various legal task capabilities.
In future, we would like to extend our work for district-level court cases, which are available in Indian languages.
Limitations
We instruction-tune our pretrained legal models on 10,763 instructions generated from Open AI GPT-3.5-Turbo,after removing the duplicates. Therefore, our instruction-tuned models may generate grammatically incoherent, and non-factual legal text generation related to Indian legal jurisdiction and constitution of India as we could not verify the instruction dataset at our end. Due to limited resources, we could not measure the degree of hallucination of our legal generative language models and also could not perform legal expert human evaluation of the responses of our legal models. We also did not anonymize the publicly available legal cases during tokenization and pretraining. We also have not developed any guardrails or preprocessing of input prompts and the outputs shown in the paper are the direct responses from the model without any rule-based post processing fact-checker or toxicity classifier. It may happen that our models may generate biased opinions, factually incorrect information as we kept the temperature value to 1.0 during text generation to allow the models to be highly creative.
Acknowledgements
We would like to dedicate the work to billions of Indians who get stuck in legal processing for decades. We hope our work can aid in legal processing by using it as an end-to-end tool. The first author also wants to dedicate his work to his beloved parents, Rita Niyogi and Malay Niyogi for their outstanding support throughout his journey.
References
- Ainslie etal. (2023)Joshua Ainslie, James Lee-Thorp, Michiel deJong, Yury Zemlyanskiy, Federico Lebrón, and Sumit Sanghai. 2023.Gqa: Training generalized multi-query transformer models from multi-head checkpoints.
- Ayansk11 (2023)Ayansk11. 2023.Legal instructions.
- Ba etal. (2016)JimmyLei Ba, JamieRyan Kiros, and GeoffreyE. Hinton. 2016.Layer normalization.
- Beltagy etal. (2020)IzBeltagy, MatthewE. Peters, and Arman Cohan. 2020.Longformer: The long-document transformer.
- Bisk etal. (2019)Yonatan Bisk, Rowan Zellers, RonanLe Bras, Jianfeng Gao, and Yejin Choi. 2019.Piqa: Reasoning about physical commonsense in natural language.
- Black etal. (2022)Sid Black, Stella Biderman, Eric Hallahan, Quentin Anthony, Leo Gao, Laurence Golding, Horace He, Connor Leahy, Kyle McDonell, Jason Phang, Michael Pieler, USVSNSai Prashanth, Shivanshu Purohit, Laria Reynolds, Jonathan Tow, Ben Wang, and Samuel Weinbach. 2022.Gpt-neox-20b: An open-source autoregressive language model.
- Chalkidis etal. (2020)Ilias Chalkidis, Manos Fergadiotis, Prodromos Malakasiotis, Nikolaos Aletras, and Ion Androutsopoulos. 2020.LEGAL-BERT: The muppets straight out of law school.In Findings of the Association for Computational Linguistics: EMNLP 2020, pages 2898–2904, Online. Association for Computational Linguistics.
- Chowdhery etal. (2024)Aakanksha Chowdhery, Sharan Narang, Jacob Devlin, Maarten Bosma, Gaurav Mishra, Adam Roberts, Paul Barham, HyungWon Chung, Charles Sutton, Sebastian Gehrmann, Parker Schuh, Kensen Shi, Sashank Tsvyashchenko, Joshua Maynez, Abhishek Rao, Parker Barnes, YiTay, Noam Shazeer, Vinodkumar Prabhakaran, Emily Reif, Nan Du, Ben Hutchinson, Reiner Pope, James Bradbury, Jacob Austin, Michael Isard, Guy Gur-Ari, Pengcheng Yin, Toju Duke, Anselm Levskaya, Sanjay Ghemawat, Sunipa Dev, Henryk Michalewski, Xavier Garcia, Vedant Misra, Kevin Robinson, Liam Fedus, Denny Zhou, Daphne Ippolito, David Luan, Hyeontaek Lim, Barret Zoph, Alexander Spiridonov, Ryan Sepassi, David Dohan, Shivani Agrawal, Mark Omernick, AndrewM. Dai, ThanumalayanSankaranarayana Pillai, Marie Pellat, Aitor Lewkowycz, Erica Moreira, Rewon Child, Oleksandr Polozov, Katherine Lee, Zongwei Zhou, Xuezhi Wang, Brennan Saeta, Mark Diaz, Orhan Firat, Michele Catasta, Jason Wei, Kathy Meier-Hellstern, Douglas Eck, Jeff Dean, Slav Petrov, and Noah Fiedel.2024.Palm: scaling language modeling with pathways.J. Mach. Learn. Res., 24(1).
- Clark etal. (2018)Peter Clark, Isaac Cowhey, Oren Etzioni, Tushar Khot, Ashish Sabharwal, Carissa Schoenick, and Oyvind Tafjord. 2018.Think you have solved question answering? try arc, the ai2 reasoning challenge.
- Colombo etal. (2024)Pierre Colombo, TelmoPessoa Pires, Malik Boudiaf, Dominic Culver, Rui Melo, Caio Corro, Andre F.T. Martins, Fabrizio Esposito, VeraLúcia Raposo, Sofia Morgado, and Michael Desa. 2024.Saullm-7b: A pioneering large language model for law.
- Cui etal. (2023)Jiaxi Cui, Zongjian Li, Yang Yan, Bohua Chen, and LiYuan. 2023.Chatlaw: Open-source legal large language model with integrated external knowledge bases.
- Fei etal. (2023)Zhiwei Fei, Xiaoyu Shen, Dawei Zhu, Fengzhe Zhou, Zhuo Han, Songyang Zhang, Kai Chen, Zongwen Shen, and Jidong Ge. 2023.Lawbench: Benchmarking legal knowledge of large language models.
- Gordon etal. (2012)Andrew Gordon, Zornitsa Kozareva, and Melissa Roemmele. 2012.SemEval-2012 task 7: Choice of plausible alternatives: An evaluation of commonsense causal reasoning.In *SEM 2012: The First Joint Conference on Lexical and Computational Semantics – Volume 1: Proceedings of the main conference and the shared task, and Volume 2: Proceedings of the Sixth International Workshop on Semantic Evaluation (SemEval 2012), pages 394–398, Montréal, Canada. Association for Computational Linguistics.
- Guha etal. (2023)Neel Guha, Julian Nyarko, DanielE. Ho, Christopher Re, Adam Chilton, Aditya Narayana, Alex Chohlas-Wood, Austin Peters, Brandon Waldon, Daniel Rockmore, Diego Zambrano, Dmitry Talisman, Enam Hoque, Faiz Surani, Frank Fagan, Galit Sarfaty, GregoryM. Dickinson, Haggai Porat, Jason Hegland, Jessica Wu, Joe Nudell, Joel Niklaus, JohnJ Nay, JonathanH. Choi, Kevin Tobia, Margaret Hagan, Megan Ma, Michael Livermore, Nikon Rasumov-Rahe, Nils Holzenberger, Noam Kolt, Peter Henderson, Sean Rehaag, Sharad Goel, Shang Gao, Spencer Williams, Sunny Gandhi, Tom Zur, Varun Iyer, and Zehua Li. 2023.Legalbench: A collaboratively built benchmark for measuring legal reasoning in large language models.In Thirty-seventh Conference on Neural Information Processing Systems Datasets and Benchmarks Track.
- Gururangan etal. (2020)Suchin Gururangan, Ana Marasović, Swabha Swayamdipta, Kyle Lo, IzBeltagy, Doug Downey, and NoahA. Smith. 2020.Don’t stop pretraining: Adapt language models to domains and tasks.In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, pages 8342–8360, Online. Association for Computational Linguistics.
- Hendrycks etal. (2021a)Dan Hendrycks, Collin Burns, Steven Basart, Andy Zou, Mantas Mazeika, Dawn Song, and Jacob Steinhardt. 2021a.Measuring massive multitask language understanding.In International Conference on Learning Representations.
- Hendrycks etal. (2021b)Dan Hendrycks, Collin Burns, Saurav Kadavath, Akul Arora, Steven Basart, Eric Tang, Dawn Song, and Jacob Steinhardt. 2021b.Measuring mathematical problem solving with the MATH dataset.In Thirty-fifth Conference on Neural Information Processing Systems Datasets and Benchmarks Track (Round 2).
- Hoffmann etal. (2022)Jordan Hoffmann, Sebastian Borgeaud, Arthur Mensch, Elena Buchatskaya, Trevor Cai, Eliza Rutherford, Diego deLasCasas, LisaAnne Hendricks, Johannes Welbl, Aidan Clark, Thomas Hennigan, Eric Noland, Katherine Millican, George vanden Driessche, Bogdan Damoc, Aurelia Guy, Simon Osindero, Karén Simonyan, Erich Elsen, Oriol Vinyals, Jack Rae, and Laurent Sifre. 2022.An empirical analysis of compute-optimal large language model training.In Advances in Neural Information Processing Systems, volume35, pages 30016–30030. Curran Associates, Inc.
- Huang etal. (2023)Quzhe Huang, Mingxu Tao, Chen Zhang, Zhenwei An, Cong Jiang, Zhibin Chen, Zirui Wu, and Yansong Feng. 2023.Lawyer llama technical report.
- Kudo and Richardson (2018)Taku Kudo and John Richardson. 2018.SentencePiece: A simple and language independent subword tokenizer and detokenizer for neural text processing.In Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing: System Demonstrations, pages 66–71, Brussels, Belgium. Association for Computational Linguistics.
- Levesque etal. (2012)HectorJ. Levesque, Ernest Davis, and Leora Morgenstern. 2012.The winograd schema challenge.In Proceedings of the Thirteenth International Conference on Principles of Knowledge Representation and Reasoning, KR’12, page 552–561. AAAI Press.
- Lin etal. (2022)Stephanie Lin, Jacob Hilton, and Owain Evans. 2022.TruthfulQA: Measuring how models mimic human falsehoods.In Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 3214–3252, Dublin, Ireland. Association for Computational Linguistics.
- Malik etal. (2021)Vijit Malik, Rishabh Sanjay, ShubhamKumar Nigam, Kripabandhu Ghosh, ShouvikKumar Guha, Arnab Bhattacharya, and Ashutosh Modi. 2021.ILDC for CJPE: Indian legal documents corpus for court judgment prediction and explanation.In Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers), pages 4046–4062, Online. Association for Computational Linguistics.
- Narang etal. (2021)Sharan Narang, HyungWon Chung, YiTay, Liam Fedus, Thibault Fevry, Michael Matena, Karishma Malkan, Noah Fiedel, Noam Shazeer, Zhenzhong Lan, Yanqi Zhou, Wei Li, Nan Ding, Jake Marcus, Adam Roberts, and Colin Raffel. 2021.Do transformer modifications transfer across implementations and applications?In Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing, pages 5758–5773, Online and Punta Cana, Dominican Republic. Association for Computational Linguistics.
- NebulaSense (2023)NebulaSense. 2023.Legal clause instructions.
- Nigam etal. (2024)Shubham Nigam, Anurag Sharma, Danush Khanna, Noel Shallum, Kripabandhu Ghosh, and Arnab Bhattacharya. 2024.Legal judgment reimagined: PredEx and the rise of intelligent AI interpretation in Indian courts.In Findings of the Association for Computational Linguistics ACL 2024, pages 4296–4315, Bangkok, Thailand and virtual meeting. Association for Computational Linguistics.
- Niklaus and Giofre (2023)Joel Niklaus and Daniele Giofre. 2023.Can we pretrain a SotA legal language model on a budget from scratch?In Proceedings of The Fourth Workshop on Simple and Efficient Natural Language Processing (SustaiNLP), pages 158–182, Toronto, Canada (Hybrid). Association for Computational Linguistics.
- Pilehvar and Camacho-Collados (2019)MohammadTaher Pilehvar and Jose Camacho-Collados. 2019.WiC: the word-in-context dataset for evaluating context-sensitive meaning representations.In Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers), pages 1267–1273, Minneapolis, Minnesota. Association for Computational Linguistics.
- Press and Wolf (2017)Ofir Press and Lior Wolf. 2017.Using the output embedding to improve language models.In Proceedings of the 15th Conference of the European Chapter of the Association for Computational Linguistics: Volume 2, Short Papers, pages 157–163, Valencia, Spain. Association for Computational Linguistics.
- Sennrich etal. (2016)Rico Sennrich, Barry Haddow, and Alexandra Birch. 2016.Neural machine translation of rare words with subword units.In Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 1715–1725, Berlin, Germany. Association for Computational Linguistics.
- Shazeer (2019)Noam Shazeer. 2019.Fast transformer decoding: One write-head is all you need.
- Shazeer (2020)Noam Shazeer. 2020.Glu variants improve transformer.
- Shukla etal. (2022)Abhay Shukla, Paheli Bhattacharya, Soham Poddar, Rajdeep Mukherjee, Kripabandhu Ghosh, Pawan Goyal, and Saptarshi Ghosh. 2022.Legal case document summarization: Extractive and abstractive methods and their evaluation.In Proceedings of the 2nd Conference of the Asia-Pacific Chapter of the Association for Computational Linguistics and the 12th International Joint Conference on Natural Language Processing (Volume 1: Long Papers), pages 1048–1064, Online only. Association for Computational Linguistics.
- Su etal. (2024)Jianlin Su, Murtadha Ahmed, YuLu, Shengfeng Pan, Wen Bo, and Yunfeng Liu. 2024.Roformer: Enhanced transformer with rotary position embedding.Neurocomput., 568(C).
- Touvron etal. (2023)Hugo Touvron, Thibaut Lavril, Gautier Izacard, Xavier Martinet, Marie-Anne Lachaux, Timothée Lacroix, Baptiste Rozière, Naman Goyal, Eric Hambro, Faisal Azhar, Aurelien Rodriguez, Armand Joulin, Edouard Grave, and Guillaume Lample. 2023.Llama: Open and efficient foundation language models.
- Vaswani etal. (2017)Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, AidanN. Gomez, Łukasz Kaiser, and Illia Polosukhin. 2017.Attention is all you need.In Proceedings of the 31st International Conference on Neural Information Processing Systems, NIPS’17, page 6000–6010, Red Hook, NY, USA. Curran Associates Inc.
- Wang etal. (2018)Alex Wang, Amanpreet Singh, Julian Michael, Felix Hill, Omer Levy, and Samuel Bowman. 2018.GLUE: A multi-task benchmark and analysis platform for natural language understanding.In Proceedings of the 2018 EMNLP Workshop BlackboxNLP: Analyzing and Interpreting Neural Networks for NLP, pages 353–355, Brussels, Belgium. Association for Computational Linguistics.
- Yang etal. (2021)Greg Yang, EdwardJ Hu, Igor Babuschkin, Szymon Sidor, Xiaodong Liu, David Farhi, Nick Ryder, Jakub Pachocki, Weizhu Chen, and Jianfeng Gao. 2021.Tuning large neural networks via zero-shot hyperparameter transfer.In Advances in Neural Information Processing Systems.
- Yao etal. (2021)Yunzhi Yao, Shaohan Huang, Wenhui Wang, LiDong, and Furu Wei. 2021.Adapt-and-distill: Developing small, fast and effective pretrained language models for domains.In Findings.
- Zhang and Sennrich (2019)Biao Zhang and Rico Sennrich. 2019.Root mean square layer normalization. Curran Associates Inc., Red Hook, NY, USA.
- Zhong etal. (2024)Wanjun Zhong, Ruixiang Cui, Yiduo Guo, Yaobo Liang, Shuai Lu, Yanlin Wang, Amin Saied, Weizhu Chen, and Nan Duan. 2024.AGIEval: A human-centric benchmark for evaluating foundation models.In Findings of the Association for Computational Linguistics: NAACL 2024, pages 2299–2314, Mexico City, Mexico. Association for Computational Linguistics.
Appendix
Appendix A Legal Tasks for Instruction Tuning
Some examples of legal tasks for instruction tuning are:
- 1.
Analyze and explain the legal reasoning behind the judgment in the given case
- 2.
Identify and summarize the key legal issues in the provided case
- 3.
Draft an argument appealing the decision of the given case
- 4.
Identify the legal precedents used in the presented case
- 5.
Draft a summary of a given Indian law or statute, outlining its purpose, main provisions, and implications
- 6.
Develop a legal strategy for a hypothetical client based on the facts of the provided case
- 7.
Draft a hypothetical dissenting opinion for the provided case
- 8.
Identify potential policy changes that could be advocated for in light of the given case
- 9.
Draft a hypothetical legal notice based on the facts of the provided case
- 10.
Discuss potential legal reforms suggested by the decision in the provided case
- 11.
Summarize the primary dissenting arguments in the provided case
- 12.
Identify areas of disagreement between judges’ opinions in the presented case
- 13.
Specify the terms of termination, including the notice period and grounds for termination
- 14.
Create a legal condition for Confidentiality Obligation for Employment Contract in Banking industry
- 15.
Add a clause to the contract that specifies the consequences of termination for convenience
- 16.
Draft a condition related to property damage for Land Lease Agreement in Agriculture industry
- 17.
Draft a summary of the given case, highlighting its purpose, main provisions, and implications
- 18.
Establish the employee’s confidentiality obligations
Appendix B Legislative Response
Legal Instruction: Discuss potential legislative responses to the decision in the given case.
Input: Indra Sawhney vs Union of India, AIR 1993 SC 477.
Response 1:
Following the reservation policy implemented by the Parliament, the legislature may consider revising the Constitution of India to require adequate representation of SCs/STs in promotions. This would provide a dedicated forture and support for affirmative action across the country.
Additionally, the legislature may introduce reservation policies to ensure that underrepresentation of socially and educationally backward classes, the exclusion of the creamy layer, the exclusion of the creamy layer, and the need for periodic review of reservation policies to ensure their alignment with constitutional principles. This would help prevent any misuse or misappropriation of reserved seats and ensure that reservation benefits reach the intended beneficiaries.
Moreover, the legislature can consider enacting specific legislation based on quantifiable data, backwardness, and economic factors. This would provide clear guidelines and measures for formulating policies that address the effectiveness and implementation of reservation systems and ensure transparency in the implementation of reservation policies.
Additionally, the legislature can work towards creating a society where every person has equal opportunity to get his name included, contributing to the promotion of masses by promoting social justice and equitable access to education. This would ensure equal opportunities for all, regardless of their caste or community.
The legislature can further explore the possibility of adopting a more comprehensive approach that considers the upliftment of these communities and promotes social justice. This could include schemes for their socio-economic progress and development, improving the collection and analysis of digital platforms and technology in the formulation and implementation of reservation policies.
It is important for the legislature to carefully consider these potential legislative responses in order to uphold the constitutional rights and equality of the SCs/STs and promote social justice and equality in India
Response 2:
Validity of reservation in promotions for SCs/STs under Articles 16(4) and 16(4A) can be challenged on the grounds of violating these constitutional provisions. However, it must be shown that the reservation in promotions affects the efficiency of administration and enables inclusivity in administration. Moreover, the legislature can consider other legitimate social factors, such as the creation of the avenues of social justice and affirmative action. These factors contribute to the overall functioning of public services and promotions. The progressive approach to affirmative action and representation in public services reaches a moment in areas where caste-based policies and regulations are inclusively considered as a criterion for determining backwardness, thereby promoting equitable access to education and opportunities. Any future reservation measures or policies that impose excessive disadvantages or disproportionate burdens on the community can be struck down if they are found to be in violation of fundamental rights.Additionally, the legislature can consider introducing a specific legislation or amendment to provide special provisions for affirmative action. This would help safeguard the efficiency and effectiveness of the law and ensure equal opportunities for all individuals.
Furthermore, the legislature can establish a comprehensive review and revision mechanism for determining the adequacy of available reservation policies. This would help identify any shortcomings or infringement on the rights of the general category candidates and address any unforeseen issues.
Overall, the legislative response to these potential legislative responses should be based on the principles of social justice, equality, and representation for SCs/STs in higher positions. It is essential to strike a balance between providing opportunities for historically disadvantaged groups and maintaining administrative efficiency. It’s further expected resources and experiences in public services would help in promoting equitable access to education and ensuring equal opportunities for all sections of society.Overall, the legislative responses to the decision in the case of Indra Sawhney vs Union of India highlight the need for a robust and equitable system of promotions to protect the rights and well-being of marginalized communities. This would help in addressing any unintended consequences or disparities that may arise from the changing needs and demands of the society. Overall, the legislative responses to the decision in the Indra Sawhney case highlight the need for a robust legal framework to address any unforeseen cases of categorization in promotions. This could involve considering the implementation of existing legislation and the improving quality of service conditions in different sectors, such as promotions, employment opportunities, and social welfare schemes. Citing the legal contextational importance of reservation in promotions for SCs/STs and upholding the principles of social justice and equality enshrined in the Constitution, ensuring equal opportunities for all citizens in various contexts and promoting inclusivity in the Indian society. Support with relevant legislative developments and policies in India and seeing the enactment of the Central Educational Institutions Reservation Act, 2006. This will help in formulating targeted policies, promoting transparency, accountability, and opening of workforms to address the underlying causes of backwardness and promoting unequal access to education. Additionally, effective implementation and monitoring of reservation policies could take into account the implementation of empowering women and ensuring their understanding of these policies.
Conclusion:
Overall, the legislative responses to the decision in the Indra Sawhney case highlight the need for a robust legal framework to address any unforeseen cases of exclusion, equality, and adequacy of various areas of social justice, and educational opportunities for the upliftment of backward classes. It also make it clear that the identification of backward classes and the need to review their existing policies and practices in line with the principles of equality and social justice. Supervising the role of the legislature in ensuring effective implementation and enforcement of reservation policies in India is crucial for the development of social justice and equality standards in all sectors. Supervising the Directorate General of Civil Provinces and NGOs for carrying out public functions, the legislature can work together with experienced and skilled independent bodies responsible for evaluating reservation policies, ensuring compliance with the law, and upholding the constitutional principles of equality and social justice.
Appendix C Key Legal Issues
Legal Instruction: Identify and summarize the key legal issues in the given case.
Input: Naresh Shridhar Mirajkar And Ors vs State Of Maharashtra And Anr (1966)
Response 1:
From the Naresh Shridhar Mirajkar And Ors vs State Of Maharashtra And Anr (1966) case revolve around the interpretation and application of constitutional provisions. The key legal issues in the case of Naresh Shridhar Mirajkar And Ors vs State Of Maharashtra And Anr (1966) are as follows:
1. Whether the High Court order prohibiting the publication of witness testimony during a trial violates Article 20(1) of the Constitution?
2. If not, whether the order causing significant harm to the witness person during the trial is addressed to a court judge?
3. If not, whether the order adversely affects the right to freedom of speech and expression under Article 19(1)(a)?
4. Whether the High Court order prohibiting publication of witness testimony infringes upon the fundamental right to freedom of speech and expression under Article 19(1)(a)?
Response 2:
For the case of Naresh Shridhar Mirajkar And Ors vs State Of Maharashtra And Anr (1966), the primary legal issues revolve around the interpretation and application of constitutional provisions related to freedom of speech and expression and the inherent powers of the High Court, respectively. The key legal principles involved are the right to a fair trial, the right to life and personal liberty under Article 21 of the Indian Constitution.
Under Article 21 of the Indian Constitution, every person has a right to a fair and unbiased hearing. This right extends to every person, and there may be exceptions to this right.
Additionally, Article 32 of the Constitution, provides for the continuation of a writ petition, which can be filed in the Supreme Court after the commencement of the Constitution. Explain the scope of these rights and the remedies available to the petitioners in this case.
Additionally, Article 32 provides a remedy for enforcing those rights. The petitioner can present his case or seek relief through a writ petition in the Supreme Court under Article 32. It is important to identify cases where similar legal issues arise from other jurisdictions or international jurisprudence involving public interests.
Overall, the case of Naresh Shridhar Mirajkar And Ors vs State Of Maharashtra And Anr (1966) is significant in the context of this specific case. It emphasizes the importance of protecting fundamental rights, such as the right to freedom of speech and expression, and accessibility to justice. The court’s decision in this case will continue to hold the door for future cases involving violations of fundamental rights and will contribute to the development of jurisprudence on this particular matter.
While the direct impact of this case on the scope and limitations of the Indian legal system is apparent, it does not explicitly discuss or provide an exhaustive list of all areas of law that can influence future cases. The decision of the Supreme Court, which balances various factors, can indirectly support the recognition of the right to a fair trial and the rights of the accused.Furthermore, the case demonstrates the power of the High Court to reach interim orders and injunctions during such proceedings. The grant of bail to the petitioner during the pendency of the appeal may have an impact on the freedom of speech and expression under Article 21 of the Indian Constitution.Overall, the case of Naresh Shridhar Mirajkar And Ors vs State Of Maharashtra And Anr (1966) is significant in the context of freedom of speech and expression and doesn’t directly impact the interpretation and application of the legal principles established in the Naresh Shridhar Mirajkar case. Lectural considerations and the right to freedom of speech and expression in the context of freedom of speech and expression may also have an impact on freedom of speech and expression. Lawyour, the right to a fair trial, has been invoked to protect the publication of private information even without the witness filing the return. The court’s remaining amenable to writ jurisdiction under Article 32 of the Constitution is not completely outcomes. The decision of the Supreme Court in the Naresh Shridhar Mirajkar case will serve as a guiding precedent for future cases involving conflicts between freedom of speech and fair trial rights. Overall, the case of Naresh Shridhar Mirajkar and Others vs State Of Maharashtra And Anr (1966) is significant in the context of freedom of speech and expression and doesn’t directly impact the interpretation of the legal principles at stake. Furthermore, the case establishes the criteria for granting bail in cases involving high treasonable circumstances. The court’s duty to protect the freedom of speech and expression under Article 21 of the Indian Constitution.
Appendix D Drafting Legal Arguments
Legal Instruction: Draft an argument appealing the decision of the given case.
Input: Naresh Shridhar Mirajkar And Ors vs State Of Maharashtra And Anr (1966)
Response 1:
Your Honor, we are here to challenge the decision of the High Court of Maharashtra in this case. The main issue in this case is whether the High Court’s order directing the publication of witness testimony infringes upon the fundamental right to freedom of speech and expression under Article 19(1)(a) of the Constitution. We urge the court to declare that the High Court does have the power to hold a trial and dispose of proceedings in the trial, which is a crucial remedy for individuals who have suffered due to the violation of their rights. Furthermore, we also demand the following:
1. Compensation for the victim’s due to the arbitrary termination of their jobs.
2. Compensation for the harm caused by the arbitrary and illegal detention of the accused.
3. Policy advocacy subjection: We contend that compensation should be paid in cases of established violations of fundamental rights. This will help prevent unnecessary hardship to the victim and ensure a fair trial.
Response 2:
Your Honor, we are here to challenge the decision of the High Court of Maharashtra in the case of Naresh Shridhar Mirajkar And Ors vs State Of Maharashtra And Anr (1966). We contend that the High Court’s decision to uphold the validity of certain provisions of the Bombay Public Trusts Act, 1950 violates their fundamental rights under Articles 25 and 26 of the Constitution.
Although the case does not directly address any issues related to the rights guaranteed under Articles 25 and 26, it provides a broader framework for interpreting and applying the relevant provisions of the Constitution to determine the validity of those provisions. The High Court’s interpretation of the constitutional principles of freedom of religion and the secularism enshrined in the Constitution is a broader question.
Furthermore, the decision in this case reinforces the importance of respecting the religious beliefs and practices of the religious community as protected under Articles 25 and 26. It highlights the need to reconsider the court’s decisions and potentially to balance the rights of these communities with the broader societal goal of their own growth.
Additionally, the decision explores the scope of judicial review in matters related to appointments, reservation policies, and the appointment of trustees or religious heads. The courts have referred to the judgment of the Supreme Court in the case of Naresh Shridhar Mirajkar And Ors vs State Of Maharashtra And Anr (1966), which emphasizes the significance of plurality in the administration of public and religious trusts.
Overall, the decision in this case contributes to the development of the legal principles at stake when challenging the validity of specific provisions like the one enacted by the ruling party in the case. Revisions for the protection of fundamental rights and the proper exercise of religious rights, if necessary, can help prevent future disputes outside of the camema structure.
Appendix E Legal Precedents
Legal Instruction: Identify the legal precedents used in the presented case.
Input:Naresh Shridhar Mirajkar And Ors vs State Of Maharashtra And Anr (1966)
Response:
From the Naresh Shridhar Mirajkar And Ors vs State Of Maharashtra And Anr (1966), the legal precedents used in the presented case of Naresh Shridhar Mirajkar And Ors vs State Of Maharashtra And Anr (1966) include: 1. The interpretation and scope of Article 21 (right to life and personal liberty) of the Constitution of India.
The court referred to its previous judgment in the Supreme Court of India, which recognized the inherent power of the High Court to prohibit the publication of court proceedings during a trial to ensure a fair trial. This judgment was relied upon by the court to uphold the right to freedom of speech and expression under Article 19(1)(a) of the Indian Constitution.
The court also considered the implications of the decision in the Naresh Shridhar Mirajkar case. It held that the High Court has inherent power to regulate its own process and issue writs for enforcing speech and expression as part of the constitutional scheme. This power is derived from fundamental rights, including the right to life and personal liberty guaranteed under Article 21. The court relied on its jurisdiction to hold that the High Court has such power, even if it indirectly contributed to the protection of detained persons’ rights.
Additionally, the case highlights the importance of balancing the right to freedom of speech and expression with other fundamental rights. It emphasizes the need to ensure a fair trial, considering the innocence of the accused, and the potential harm to his unheard. The court’s reasoning is based on the concept of administering justice according to the principles established in the landmark case.
Furthermore, the court referred to its previous judgement in the Supreme Court of India case, where the court exercised restraint in interfering with the fair administration of justice. This precedent contributes to the evolution of Indian jurisprudence regarding the powers of the High Court and the limitations on its own powers
Appendix F Raising Hypothetical Questions
Legal Instruction: Develop a hypothetical question that could be raised in parliament based on the provided case.
Input: Naresh Shridhar Mirajkar And Ors vs State Of Maharashtra And Anr (1966).
Response:
What steps is the government taking to ensure that public functions and duties are performed by stakeholders, as highlighted in the case of Naresh Shridhar Mirajkar And Ors vs State Of Maharashtra And Anr (1966)? How does the government plan to address any challenges to the stability or continuity of a criminal charge prone to significantly offend constitutional rights? Was the government considering any reforms to strengthen the client’s case-related to the principles of fairness and equality in education? What measures are being considered to hold the state accountable for any arbitrary actions or discriminatory practices in prison administration? What measures are being considered to hold the state accountable for any violations of fundamental rights and ensure that individuals are treated fairly and equally in accordance with the law?
Additionally, the government’s decision to lay down guidelines and regulations for the prevention of sexual harassment at the workplace states raises questions about the role of employers in creating a safety system and promoting gender equality. The government should establish an independent regulatory body to oversee the workplace and provide necessary support and skill development opportunities to its employees, including access to legal documentation, to monitor implementation procedures, and attend all sectors in the organization. This will help in compliance with relevant laws and regulations while maintaining the integrity and effectiveness of the journalistic process.
It has come to our firm positive that the issue of addressing this issue is predominantly and the government’s decision becomes crucial. It may lead to an increase in addressing grievances related to sexual harassment, the establishment of a more safe and conducive working environment. Additionally, it is essential for the government to take timely action to reform the conditions of arrest, detention, and other crimes incidents whenever required.
Appendix G Potential Policy Changes
Legal Instruction: Identify potential policy changes that could be advocated for in light of the given case.
Input: Naresh Shridhar Mirajkar And Ors vs State Of Maharashtra And Anr (1966)
Response:
Clear guidelines for eviction and the definition of ’law’ under Article 21 of the Constitution: The case highlights the need for potential policy changes to address environmental issues and protect the rights of pavement dwellers and prevent violations of their fundamental rights. Some potential policy changes that could be advocated for in light of this case are:
1. Regulation of Preventing Wildlife Protection Control: The government should take immediate steps to hold elected organizations accountable for their inability to file complaints related toten. Such actions may include schemes for the improvement of workforce, mental health, moral support, and other stakeholder responsible persons in occupation.
2. Training and sensitization: The case also raises the issue of poverty and lack of access to basic necessities. This could involve training programs for pavement dwellers who are driven to cities by poverty and lack of rural jobs. Such a policy should outline the importance of addressing these issues through proper identification and identification processes.
3. Strengthening Legal Reforms: The government should closely monitor the infrastructure requirements and teaching mechanisms for pavement dwellers. This can be done on monthly basis, reducing timber-related precautions, and promoting inclusivity in the workforce.
4. Legal Reforms: The government should introduce legislation specifically to amend existing laws to provide legal backing for pavement dwellers. This would help strike a balance between protecting fundamental rights, ensuring access to basic necessities, and ensuring justice for marginalized communities.
5. Collaboration with Civil Appeals and Companion cases: The government should establish a coordination between various sporting bodies and their institutions, providing financial assistance, and ensuring access to infrast.
Appendix H Legal Strategy
Legal Instruction: Formulate a potential legal strategy for a client seeking to avoid a similar issue as in the given case.
Input: The case Indra Sawhney vs Union of India, AIR 1993 SC 477
Response 1:
For a client who is a member of a statutory corporation or a state-appellorees to implement the reservation policy for backward classes in promotions. The client should ensure that their representation is based on merit and not caste as a whole. They can utilize the unique nature of their duties and can make necessary adjustments to society.
Specifically, if the client belongs to a socially disadvantaged group, they may consider alternative measures to uplift the most marginalized sections of society. This includes addressing the underrepresentation of other categories, providing a clearer definition of backwardness, and securing social justice.
By implementing these strategies, the client can proactively mitigate the risk of promoting social justice and equality in promotions. They should adopt fair practices and non-discrimination to avoid any constitutional challenges or emphasize the importance of affirmative action.
Additionally, the client should consult with a qualified lawyer specializing in employment and constitutional law to understand the specific circumstances and needs of each particular industry or employer. This assists in managing the available tailored areas.
By implementing these strategies, the client can minimize the risk of promoting social justice and equality in promotions and ensure a fair and inclusive education system in their country. It can also be beneficial to ensure equal opportunities for all, regardless of their backwardness and promoting equal opportunities.Considering the client’s situation and suitability for monitoring and adherence to constitutional principles, it is crucial to establish a transparent and accountable analysis of the constitutional provisions involved. This includes striking a balance between protecting the rights of different groups and ensuring equal opportunities for all.
Lastly, the client should consult with a qualified lawyer with expertise in constitutional law to provide effective remedies for their grievances and relief.
Response 2:
For a client who is a general category candidate seeking admission to a professional or technical college and has served in the capacity of a scholarly designated society, it is important to avoid any potential issues arising in this case. The client should ensure that their applications are made in a fair and non-discriminatory manner, following the merit-based selection process. This includes providing equal opportunities for all students and maintaining their family libratics.
The client should establish that they have been appointed regularly based on their eligibility for promotion as per the quota reserved for candidates from other communities. This will help in demonstrating whether the selection process is arbitrary or malafide. It should also establish that they have no right to question the special qualifications of candidates with lesser merit compared to those without promotion.
Additionally, the client should maintain a balance between their right to equality under Articles 14 and 16(1) of the Constitution. They should recognize that they are not disparied within a reasonable time frame, leading to discrimination against them based on their abilities rather than their right to equal opportunity.
It is important to our client to consult with an experienced lawyer who specializes in constitutional law and employment matters. The lawyer can provide personalized guidance based on the specific circumstances of each case, including the nature of the appointment and animal labor relationship.Overall, the legal strategy should focus on advising our client to ensure that their applications are made in a fair and non-discriminatory manner, following the merit criteria and the role of the lawyer.Advocating for mitigation measures to ensure compliance with the court’s decision can help create a situation where you navigate the implications of your awarding professional responsibilities with greater clarity and effectiveness.
Appendix I Legal Notice
Legal Instruction: Draft a hypothetical legal notice based on the facts of the provided case.
Input: K. S. Ramamurthi Reddiar vs The Chief Commissioner, Pondicherry & Anr., AIR 1963 SC 1464
Response 1:
“To”: Legal Action if your legal notice revolves around the interpretation and application of the Constitution of India, specifically Article 12 of the Constitution. The Supreme Court’s ruling in the case of K. S. Ramamurthi Reddiar vs The Chief Commissioner, Pondicherry & Anr. emphasizes the importance of following an intelligible principle that identifies specific areas of law, consensus-wise, and considering multiple factorsologies. Therefore, it has significant implications for the interpretation and application of Article 12 and the extent of jurisdiction under Article 12.
Based on this legal notice, it is crucial for your attorney to file this legal notice, specifically mentioning the grounds on which the petitioner is being considered a citizen. It is essential to ensure that the relevant authorities conductively implementing the order and exercising their powers based on established legal principles.
We should explore other similar cases involving the interpretation and application of Article 12 and the extent of jurisdiction under Article 226 in the context of writ jurisdiction. This legal principle can be applied to other cases involving the maintainability of writ petitions under Article 32 against final judgments of the Supreme Court after the dismissal of a review petition.
For example, in cases where the jurisdiction of the Supreme Court to entertain a writ petition under Article 32 is challenged, any matter which the Supreme Court can consider may be subject to judicial scrutiny in the interest of justice. This legal argument can argue that there are no areas of law other than jurisdiction, legal authority, and potential misuse of such matters.
We should further contend that the Supreme Court’s previous order in the Supreme Court is a final prone to reconsideration and cannot be questioned in court. This salarare can be invoked to challenge the correctness of the earlier order and seek a rectifying of the previous Order passed by the Supreme Court after the dismissal of a review petition.
If your legal attention is drawn to this legal issue, it may be advisable to assert that we will be able to effectively moonsoon review petitions until the legality and propriety of the earlier order is reconsidered.
We should continue the legal process and review our findings. This legal process can serve as a deterring instrument, creating a precedent in future cases involving similar issues and making necessary adjustments based on justice, equity, and efficiency.Additionally, the case highlights the limited scope of judicial review in cases involving the interpretation and application of Article 12. The Supreme Court’s inherent powers under Article 12 restrict the maintainability of writ petitions under Article 32 against final judgments of the Supreme Court after the dismissal of a review petition. This restriction can serve as a safeguard against arbitrary decisions and ensure that individuals who have suffered from wrongful incarceration are adequately compensated.
Overall, the legal streamlining needs to demonstrate the Supreme Court’s commitment to upholding constitutional rights, ensuring a just and equitable society.
Response 2:
Filed: We hereby write to bring to your attention that the Law Officiators in Pondicherry are not just above the inferior standards of the Constitution. This will help preserve the independence and integrity of the judiciary in India.
We urge the government and legislature to review and reconsider the application of these constitutional provisions and continue to apply the law. It is crucial for courts to stay updated on any changes or developments in the legal landscape.
It has come to our attention that certain key amendments have been made by the Parliament and the new safeguards provided by the Constitution have been relaxed. However, we contend that they are not in violation of Articles 14, 19, and 21 of the Constitution.
Furthermore, we urge the government to strengthen the protection of prisoners’ rights from being violated and provide immediate effect to their life. It is our contention that the court has opened the door for lack of clarity and urgency regarding the interpretation of these constitutional provisions.
We would like to draw your attention to the evolving nature of the case and the need it decided to serve as a benchmark for any further discussions and determinations in the field. Law officiators’s primary focus will be required to consider whether the restrictions placed on freedom of speech and expression under Article 19(1)(a) are in line with the evolving cultural landscape.
Additionally, we urge the government and legislature to review and reconsider the implementation of the provided law by ensuring the stability and continuity of the judicial system. It is essential to strike a balance between protecting fundamental rights and maintaining the integrity of the Constitution that shapes the course of action.
We also call upon the government and legislature to carefully consider the implementation of existing laws and their international experts for transitional legislation in the context of their enforcement. If necessary, explore all remedies and compensation for victimizations
Appendix J Legal Petition
Legal Instruction: Draft a hypothetical legal petition based on the provided case.
Input: K. S. Ramamurthi Reddiar vs The Chief Commissioner, Pondicherry & Anr., AIR 1963 SC 1464
Response 1:
{ “To”: “The Hon’s Name”}
Subject: “Petitioner’s Contact In the case of K. S. Ramamurthi Reddiar vs The Chief Commissioner, Pondicherry & Anr. I am concerned about the scope of Article 12 of the Constitution of India and its application to the Chief Commissioner of Pondicherry, which deal with the transfer of the petitioner, in this case. The petitioner, K.S. Ramamurthi Reddiar, challenged the order of his removal from the post of Chief Commissioner, Pondicherry, who was appointed as an officiating Commissioner. The Chief Commissioner, in his capacity as an appellate authority under the Motor Vehicles Act, reviewed the original decision and issued orders transferring several other Judiciaryges. The petitioner filed writ petition under Article 32 of the Constitution, asserting that the transfer of the petitioner from the post of Chief Commissioner, Pondicherry, was valid. Clearly, the petitioner has experience in challenging the transfers and the Chief Commissioner of Pondicherry, who is acting as such, is job-consens with various stakeholders and is acting as the appellate authority in this case. The Supreme Court, in its judgment, affirmed the petitioner’s position as Chief Commissioner, and directed him to rectify the situation amending the current system of justice. Merely relying on this judgment, deeply, can be of no avail to the petitioner. As a result, it is important to note that the petitioner’s success in this petition will depend on the chances of reverse and uphold the rights of all parties involved. It is always advisable to seek redressal for any misconduct or misappropriation of fundamental rights.
Additionally, the petitioner’s right to freedom of speech and expression extends to the Chief Commissioner’s substantive right to speak and express his opinions. This means that even these rights are protected under the Constitution. Denying litigation against this issue will positively harm the petitioner’s dignity and rights guaranteed under the Constitution.
Furthermore, the petitioner’s case has highlighted the need for a comprehensive petition under Article 32 of the Constitution, seeking relief against the cancellation of his officiation. The court’s previous order granting reservation in promotions, specifically the rank of Lieutenant Governor, was upheld in subsequent cases, may confirm the petitioner’s position in this specific case. This demonstrates the importance of protecting the rights of the petitioner and ensuring justice is served.
Overall, the petitioner, K.S. Ramamurthi Reddiar, has a key position in the Indian legal system and has presented his case as a constitutional functionary when the matter is heard by a bench of the Supreme Court. These facts and circumstances can provide additional guidance and prevent any potential miscarriage of justice.
Overall, the case of K.S. Ramamurthi Reddiar vs The Chief Commissioner, Pondicherry & Anr. is significant in the context of the specific case dealt with by the Supreme Court of India and can serve as a persuasive authority and contributes to the interpretation and application of the relevant constitutional principles in this area. Lawyncc stems from the landmark case of K.S. Ramamurthi Reddiar vs The Chief Commissioner, Pondicherry & Anr. and reinforces the importance of protecting fundamental rights and ensuring equality of all individuals in the face of the Constitution. Present a hypothetical legal petition based on the given case, outcome the final verdict of the petitioner’s rights and the provision of the Constitution, supporting the petitioner’s position.
We would like to emphasize the need for a comprehensive reconsideration of the petitioner’s claims and the approach of the judiciary in addressing societal problems. This approach ensures that the interpretation and application of constitutional provisions related to equality and non-discrimination are guided by principles of fairness and equality.
Response 2:
“To”: [Name of the petitioner’s Name]
Subject: [Name of the petitioner’s trade or business by the respondent; ”[Your Name]
We, the petitioner in the above-mentioned case, hereby submit this petition under Article 32 of the Constitution of India to this Hon’ble Court challenging the validity of the order passed by the Chief Commissioner of Pondicherry (Appellant for 1963). This petition seeks to challenge the validity of the order, dated 1st July 1963, and the subsequent Chief Commissioner’s order, dated 2nd September 1964, as affirming the petitioner’s right to freedom of speech and expression under Article 19(1)(a) of the Constitution.PRAYER:
In view of the facts & circumstances described above, it is most respectfully prayed that this Hon’ble Court may be pleased to:
a) Issue a writ, order or direction in the nature of certiorari quashing the order passed by the Chief Commissioner of Pondicherry (Appellant for 1963).The petitioner is aggrieved by the decision of the Chief Commissioner of Pondicherry, who upheld the constitutional validity of the order.
b) Grant any other relief or remedy that this Hon’ble Court deems fit and just in the circumstances of the case.
c) admit this petition under Article 32 of the Constitution of India. name][Additionally, the petitioner’s case highlights the principle of judicial review in this case. The Chief Commissioner, Pondicherry, had no previous criminal record under the control of the Government of India and was not subject to Article 12. This circumstance believes the petitioner’s claim for relief cannot be entertained as it violates the petitioner’s fundamental rights under Articles 19 and 21.The petitioner’s case is primarily based on the interpretation of Article 12 and Article 21, which guarantees the right to life and personal liberty. However, the petitioner’s case does not directly impact their right to life and personal liberty, and it is essential to consider the contextual background and circumstances of the case.Firstly, the petitioner’s case does not sufficiently consider the grounds for granting the relief given in the Chief Commissioner’s report. The fact that the original grant has been rejected without any consideration or dissenting opinion demonstrates the importance of considering the merits of the case and the court’s reasoning.
Overall, this petition under Article 32 is based on the unconstitutionality of the impugned order and the subsequent administration of the law. It is important to consider the specific facts and circumstances of the case, as well as the approach taken by the Chief Commissioner, in making the application of the critical date of the order. Continuously assess the procedural flaws and violations of the petitioner’s fundamental rights and ensure compliance with the court’s orders.Additionally, the petitioner’s right to a fair trial is also established. The court’s refusal to issue a writ, order or direction affecting the violation of fundamental rights under Article 21, despite the severity of the sentence, does not automatically entitle the court to interfere with the discretion exercised by the Chief Commissioner. This aspect of the case is delicately discussed in the petition and does not indicate any specific legal issues or challenges arise.
Lastly, the application of the critical language used by the petitioner for its resolution is not clear and specific details of the specific circumstances of the case. Each case will be decided based on its specific facts and circumstances, including the nature of the violation of fundamental rights, the court’s order, and the specific relief sought.
Overall, this petition under Article 32 is a constitutional remedy to enforce the fundamental rights of individual citizens and hold the petitioner’s interests ineffectively discharging their legitimate obligations towards the law enforcement agencies. verification - We appreciate the court’s reliance on these arguments and the specific relief sought in the Chief Commissioner of Pondicherry. After weighing the court’s judgment and directing the petitioner to rectify the situation until reconsidered. If necessary, the Chief Commissioner of Pondicherry, after considering all relevant documents, and after considering all relevant factors, may pass an interim order or order restoring the petitioner’s right to freedom of speech and expression. This order likely encourage the petitioner’s participation and cooperation throughout the proceedings.”
Introduction of the Case: [Your Name][Your Advocate’s Name][Your Advocate’s Designation][Current Date] [Present the petitioner’s case and the significance of filing the Case Organization’s Name][Conclusion]The petitioner has a longstanding career and experience in communicating with the new Ministry of Commerce and Industries to protect the interests of the Indian National Standards Ltd., The petitioner’s philosophy and perception presented her case throughout the legal proceedings. It is crucial to establish an influence of this petition on the grounds for which the petitioner’s claim for compensation can be made.
If the new Ministry is not involved in the decision, we would like to bring to the attention of this Hon’ble Court the potential background and circumstances of the case, including the shortcomings noted and the specific circumstances of the case.
Given the facts and circumstances of the case, we would argue that this Hon’ble Court is more concerned about the merits of the case and the limited scope of its reconsideration in this case. [Date][Specify the specific relief sought in the Chief Commissioner of Pondicherry, and specify the relief sought in the Chief Commissioner’s report.[Law Firm’s Name][Date][Law Firm’s Name][Law Firm’s Contact Details][Specify the specific relief sought in the Chief Commissioner of Pondicherry case & continuously assess the petitioner’s grievances and seek appropriate remedies.[Date] Specify the specific relief sought in the Chief Commissioner of Pondicherry case & continuously assess the petitioner’s eligibility for relief in this case.[Law Firm’s Name][Law Firm’s Designation][Law Firm’s Contact Details]
Versus
[Undertakings of the Case]
The petitioner, K.S. Ramamurthi Reddiar, a resident of Pondicherry, was appointed as an officiator of the Union of India. The key legal issues identified in the Chartnership of this case are as follows:
1. Whether the petitioner’s appointment as a Deputy Chief Commissioner, as confirmed by the Chief Commissioner, violates the petitioner’s fundamental rights under Articles 14, 19, and 21 of the Constitution?
2. Whether the Supreme Court can review the appointment of a Deputy Chief Commissioner in a similar manner from another legal issue?”. Whether the Council for Scientific purposes, which has been declared invalid by the Supreme Court, should be considered a ’State’ under Article 12¿‘ [Your Name]”[Date]”, [Email Address][Date]”, [Law Firm’s Name]”, [Your Address]”, [Email Address]”, [Date]”, [Law Firm] Date]”, [Date]”, [Permit Address]”, [State the petitioner’s Name]”, [Specify the cause of action, etc., [Specify the cause of action, etc.”
Appendix K Legal Advice
Legal Instruction: Draft a hypothetical legal advice for a client based on the outcome of the provided case.
Input: K. S. Ramamurthi Reddiar vs The Chief Commissioner, Pondicherry & Anr., AIR 1963 SC 1464
Response 1:
Formerous Title: The case of K. S. Ramamurthi Reddiar vs The Chief Commissioner, Pondicherry & Anr. is a significant constitutional law perspective and raises important issues related to the interpretation of Article 12 of the Constitution of India. The key legal principles involved in this case are as follows:
1. Article 12 defines the term “State” for the purposes of Part III of the Constitution, which deals with Fundamental Rights. The question before the Supreme Court was whether the Chief Commissioner of Pondicherry can be considered an authority under the control of the Government of India under Article 12.
2. Article 226 empowers the High Courts to issue writs for their enforcement. The key provisions of Article 226 are the powers of the Supreme Court under article 136, the power to issue writs for the enforcement of fundamental rights and the control of the Government.
3. Article 142 directs the State to pay the minimum wages of foreign employees. The petitioner, being a Deputy Chief Commissioner of Pondicherry, argued that his salary should be fixed according to the nature of the job, irrespective of the caste or past of the person. He contended that he has been wrongfully terminated from Indian service without any valid reason or justification.
4. The Supreme Court, in its landmark judgment, held that the right to equality under Article 14 is a fundamental right under Article 14 but not Article 12. The Court ruled that compensation can be awarded for the deprivation of fundamental rights until the twenty-sixth day, i.e., 15 years of service.
The Court further emphasized that the right to equality applies even to contractual terms between State entities and private parties if found constitutionally justified, just as it applies equally to all citizens.
In summary, the case of K. S. Ramamurthi Reddiar vs The Chief Commissioner, Pondicherry & Anr. is a critical precedent and raises important questions regarding the interpretation of Article 12 and the scope of administrative decisions taken in similar circumstances.
Best regards,
[Your Name][Your Designation][Central Jurisdictional Number][DATE: This judgment clarifies the scope of administrative decisions and their implications for administrative decision-making. It establishes that the determination of a labor dispute by a local authority is unconstitutional and should be struck down as invalid.
Considering the Supreme Court’s ruling on the dismissal of a review petition, the court held that the Supreme Court should exercise its inherent powers to prevent abuse of judicial process and ensure justice. It clarified that there is no automatic absorption of statutory benefits, particularly those directly related to foreign national academic performance.
Response 2:
Formal application of Honorable Supreme Court’s decision in the case of K. S. Ramamurthi Reddiar vs The Chief Commissioner, Pondicherry & Anr. involves a Deputy Chief Commissioner of Pondicherry, specifically the jurisdiction of the Members of the Administrative Tribunal, as mentioned in Article 123 of the Constitution. This means that the transferred posts of all Civil Appeals and Class IVs are transferred to posts of Civil Appeal Nos. 1864- 1865 of 1963. The relevant provisions of the Constitution may include reference to the Supreme Court’s previous order and refer the case to a larger 7-judge bench for reconsideration. Explore alternative legal remedies, such as filing a review petition under Article 137 or withdrawing an earlier order, suitability of judicial orders and policies lacking details of the facts and law involved. If the transferred posts are in line with the principles of justice and equality, following these principles may be adopted in a detailed manner.Overall, the present case serves as a reminder to the functioning of the administration and the transfer of civil appeals from one judgment to another. It ensures transparency and fairness in the functioning of the administration, allowing citizens to continue to have control over their powers and positions.Furthermore, the case highlights the significance of both constitutional and statutory provisions in upholding the principle of separation of powers. These provisions remain essential for future cases involving similar issues and may lead to judicial decisions, potentially leading towards past matters where there are significant changes or adaptational actions. For instance, the transfer of a senior-state party toe avenue for seeking relief in the interest of public welfare, if applicable. This case has implications for the rights of employees and the responsibility of the judiciary in safeguarding individual rights. Lawyer’s legal advisor:
Overall, the present case serves as a reminder to the functioning of the administration, strengthening the protection of fundamental rights, and promoting transparency and fairness in the functioning of the administration. Collective reforms, effectively addressing the gaps and potential threats posed by various sectors, need to address the systematic issues faced by civil society groups and promote a more just and equitable society.While the specific circumstances and arguments put forth by both parties can tailor, the present case demonstrates the potential benefits and reinforcement of many of the feasiest situation surrounding privacy and dignity in the country.Overall, the present case serves as a reminder to the functioning of the administration and the promotion of civil liberties, particularly in terms of securing equality and fairness. The present case serves as a precedent for future cases involving conflicting judgments and rights related to important aspects of the administrative law.
Appendix L Legal Contract Condition
Legal Instruction: Create a legal contract condition regarding academic performance for Student Enrollment Agreement in Education industry.
Response:
YEAR FOR PROVISIONS by Laws. The Institution may terminate this Student Enrollment Agreement immediately upon written notice to the students if the student breaches any material term or condition of this Agreement, including but not limited to labor disputes, violence of applicable laws, or regulations, or violation of applicable laws. In the event of termination, the student shall be responsible for paying all amounts due and owing to the student up to the date of termination.The student shall also bear all costs associated with finalizing and entering into this agreement, including reasonable attorneys’ fees and court costs.The student agrees to comply with this academic performance condition, including but not limited to academic integrity and access to confidential information. The rejection of this academic performance request shall not constitute a waiver of any other academic performance expectations, unless such waiver is obtained prior to the intended purpose.
Upon termination of this Agreement, the student shall be released from all obligations under this Agreement and shall provide written notice to the study personnel, including a written statement of the reason for termination, and the name and address of the proposed student.If the student fails to notify the Sponsor of any unauthorized activities within thirty (30) days of receipt of an invoice from the Sponsor, then the Sponsor shall have the right to suspend the student’s life insurance coverage, including but not limited to confidentiality and non-disclosure. If the student fails to remedy this obligation, then the study personnever may pursue additional or different remedies as deemed necessary.Please note that the above conditions are important for both parties in order to ensure that the offending provisions are legally enforceable and protect the interests of the study.If the student breaches these conditions, then the Sponsor may pursue any other remedy available to it, at law or in equity
Appendix M Legal Exhibit
Legal Instruction: Create a Legal Exhibit for Work Schedule for Tenant.
Response:
“Work Schedule”:
The Tenant shall perform the services as follows:
1. Performance Scheme: The Tenant shall perform all of the following services during the [insert number] days
2. Facility Schedule: The Tenant shall perform the services on the following schedule:
- formula]
- Year: — for a period of [insert number] days]- ] for a period of [insert number] days
3. Disability Schedule
The above defects are not cured within [insert number] days of the due date.4. Quality Assurance: If the Tenant fails to meet any of the above defects, the defects shall be deemed accepted by the Tenant and the Landlord may terminate this Agreement immediately upon giving written notice to the Tenant.
5. Change Orders: The Tenant shall make all changes to the defects caused by the Tenant’s negligence or misuse of the defects. The Tenant shall return all defects in accordance with the new plan.
6. Participation: After the completion of the tenure, the Tenant shall pay a deposit of [insert amount] to the Landlord’s account. The deposit shall pass on to the Tenant at the end of the lease term.
7. Modification: Any modifications to the milestones or procedures shall be made in writing and signed by both parties.8. Survival: The provisions of this Exhibit shall survive the termination or expiration of this Agreement.9. Termination: Any termination or expiration of this Agreement shall not relieve the Tenant of its obligation to perform any additional services or take any action that would cause another Tenant’s breach of this Agreement.
Appendix N Legal Contract Cause
Legal Instruction: Create a legal contract clause for the interest rate provision for Credit Agreement in Banking and Finance industry.
Response:
“Transactions Clause”:
The interest rate applicable to this Credit Agreement shall be determined by the Lender and shall be based on a percentage of the total loan amount, less any applicable amounts, as determined by the Lender. The interest rate shall be calculated on the outstanding principal balance of the loan, less any applicable amounts owed to the Borrower.The interest rate shall apply to all loans and deposits owed to the Lender by the Borrower. If the Lender is unable to recover the loan amount due under this Credit Agreement due to non-performance or breach of the agreement, the interest rate shall be increased by a equal amount to the full amount due for the period between the date of the loan and the date of signing the transaction.This provision shall survive the termination or expiration of this Credit Agreement.The interest rate applicable to this Credit Agreement shall be reduced to [insert number of days] days from the anniversary date of this Credit Agreement.This Credit Agreement may not be modified except in writing signed by both partiesIn the event that the Lender is unable to recover the loan amount due under this Credit Agreement within [insert number of days] days of the due date, the Borrower shall have the right to prepare and return all deposits made by it in its current condition, and retain any deposits made by the Borrower as liquidated damages.The Lender reserves the right to adjust the interest rate applicable to the outstanding principal balance of the loan at any time upon request. If no such adjustment is made, the Borrower shall be entitled to payment of interest at a rate equal to the annuity rate prevailing on the date of payment.The Lender shall have the right to suspend or terminate the loan until the Lender has paid all amounts due under this Credit Agreement.
Appendix O Financial and Security Clause
Legal Instruction:Create a financial and security clause for the employment contract for Employment Contract in Technology industry.
Response:
Financial Name and Security Clause:
1. Salary: The Employee shall receive a base salary of [insert amount] per year, payable in [insert frequency] installments.
2. Bonus: The Employee shall be eligible to receive an annual bonus based on performance and market research. The bonus amount shall be determined by the Company in its sole discretion.
3. Benefits: The Employee shall be entitled to participate in any employee benefit plans or programs that are made available to similarly situated employees of the Company.
4. Paid Time Off: The Employee shall be entitled to paid time off for vacation, sick leave, and personal days, in accordance with applicable law and company policies.
5. Retirement Plan: The Employee shall be eligible to participate in the Company’s retirement plan, which includes the employee’s contributions up to a certain percentage.
6. Paid Time Off: The Employee shall be entitled to [insert number] weeks of paid vacation per year, which shall accrue upon completion of the Company’s set-off period.
7. Paid Time Off: The Employee shall be entitled to [insert number] weeks of paid vacation per year, which shall accrue on a [insert basis] basis.
8. Benefits: The Employee shall be entitled to participate in any employee benefit plans or programs that have been made available to similarly situated employees of the Company, subject to the terms and conditions of such plans or programs.
9. Mitigation: The Employee shall cooperate with the Company in the process of ensuring that the benefits packaged by the Company is being used in competition with similar Company products.
Appendix P Intellectual Property Rights
Legal Instruction:Create a legal contract clause related to the assignment of intellectual property rights for Assignment Agreement in Intellectual Property industry.
Response:
“Intellectual Property Rights Clause”:
1. Each Party hereby assigns, transfers, and conveys to the other Party all of its right, title, and interest in and to its creatures, equipment, and materials developed, including, but not limited to, patents, trademarks, copyrights, trade secrets, and any other proprietary rights, and all transactions therein provided by applicable law or regulation that by their nature are an integral part of this Agreement.
2. The Assignor represents and warrants that it is the sole owner of all rights in and to its own proprietary and confidential information, trade secrets, and any other proprietary rights, and that any such rights have been acquired pursuant to a court order or any other legal process.
3. The Assignor agrees to execute any and all documents necessary to transfer ownership of such rights to the other Party in connection with the assignment of intellectual property rights.
4. The Assignor shall indemnify and hold harmless the Assignee from any and all claims, damages, and liabilities arising out of any breach of this Agreement by the Assignee.
5. This Assignment Agreement shall be binding upon and inure to the benefit of the parties hereto and their respective successors and assigns.
6. This Assignment Agreement shall be binding upon and inure to the benefit of the parties hereto and their respective successors and assigns.
7. This Assignment Agreement shall be governed by and construed in accordance with the laws of the state of [insert state], without giving effect to any choice of law or conflict of law provisions.8. This Assignment Agreement constitutes the entire agreement between the parties with respect to the assignment of intellectual property rights under this Agreement and supersedes all prior negotiations, understandings, and agreements between the parties.
Appendix Q Non-Disclosure Agreement
Legal Instruction::Include a Non-Disclosure Agreement (NDA) clause.
Response:
“Confidentiality and Non-Disclosure Agreement (NDA) Clause”:
The parties acknowledge that in the course of their business relationship, each party may disclose to the other certain confidential information related to its business relationship. For purposes of this Agreement, “Confidential Information” means any information disclosed by one party to the other party, either directly or indirectly, in writing, orally, or by drafts, destructive resources, or that is marked as confidential or that should reasonably be understood to be confidential given the nature of the information and the circumstances of disclosure. Confidential Information includes, but is not limited to, trade secrets, customer lists, financial information, product designs, technical specifications, marketing plans, and any other information that is not generally known to the public.The parties acknowledge that in no event can the Receiving Party disclose any confidential or proprietary information related to its business relationship or any of its affiliates, without the prior written consent of the Disclosing Party.
If the Receiving Party is required by law to disclose any confidential or proprietary information, it shall provide the Disclosing Party with prompt notice of such requirement so that the Disclosing Party may seek a protective order or other appropriate remedy.This Agreement shall be binding upon the parties and their respective successors and assigns.This Agreement shall remain in effect for a period of two (2) years from the date of execution unless terminated earlier by mutual agreement of the parties in writing.Upon termination of this Agreement, the Receiving Party shall either return all Confidential Information to the Disclosing Party or certify in writing its identification of the Confidential Information to the Disclosing Party.
Appendix R Legal Tenant Contract
Legal Instruction: Draft a legal tenant contract.
Response:
“Contract Law”: This provision should specify the legal terms, jurisdiction, and scope of the agreement. It should include any provision that creates the lease agreement between the landlord and tenant, such as a written agreement signed by both parties. These provisions should outline the types of cases where the tenant may terminate the lease agreement.This provision should also outline the process for resolving any disputes related to the lease agreement and the labeling requirements of the landlord.It is important to carefully define the terms of the agreement to avoid any confusion or disputes in the future. It is always recommended to consult with a legal professional to ensure that the provision is appropriate for the specific situation and the specific circumstances surrounding the termination of the lease agreement.It is recommended to consult with a legal professional to ensure that the provision is appropriate for the specific situation and the specific requirements of the parties involved.These provisions should be reviewed annually by the lender and agreed upon by both the parties in advance and by an implementation method, as well as the specific requirements and guidelines of each party in the event of termination.
The renewed lease agreement should clearly state the status of the lease agreement and the terms of the agreement. The terms and conditions of the lease agreement will be agreed upon in advance and in writing.This provision is intended to cover the lengthy and geographic area where the lease agreement is located.This provision should outline the terms and conditions of the lease agreement, such as the manner of compensation, the method of payment, and any applicable penalties or fees.If this provision is found to be unenforceable, the following provisions may be included in the lease agreement;
1. Tenant Default: If one party fails to pay the rent or other monthly rent due under this agreement within a reasonable time period, the other party may terminate the lease agreement.
2. Termination: Upon termination of this agreement, the lessee shall immediately vacate the premises and return possession of the premises in good condition.
Appendix S Legal Question-Answering
S.1 What are the fundamental rights guaranteed under the Indian Constitution?
Response:
Yes, fundamental rights can include freedom of speech and expression, according to educational and public policy, online entities, with fundamental rights, as mentioned in Articles 134, 190, 219, and 228, have the potential to disallow any religious or secular authority that could block public order, natural justice, and encroachment. Additionally, the principles laid down in the case of M.Raman Shetty v. Shri Ram Murtaza Mia (1979) have significant implications for the protection of individual freedoms and the government’s ability to regulate public spaces.areThe Constitution of India ensures that every individual has the right to practice their religion upon their own personal beliefs and prejudices. This promotes greater respect, aim to foster trust and confidence in the judiciary, and helps maintain society integrity. Additionally, the Kesavananda Bharati case has implications for the maintenance of social harmony and the promotion of gender equality. By recognizing the right to work in an environment free from sexual harassment, the court has strengthened the scope of fundamental rights, but affirmed the concept of the state’s power to enact laws for giving effect to its policy.Its societal goals, such as the right to life and liberty, are essential for the development of a just and equitable society.It’s crucial fortimely urgent urgent steps, such as curbing the atropological discoursets during international journey philosophy, raising awareness about ways of working towards pastor-overeans.
These steps aim to ensure the protection of individual data, prevent misuse, and address social issues like the characterization of women’s self-identified gender as crucial forms of gender adversity.
S.2 What are the main provisions of the Insolvency and Bankruptcy Code, 2016?
Response:
Under Section 637 of the Insolvency and Bankruptcy Code, 2016, in cases of insolvency and bankruptcy cases, all persons interested in a liquidy or insolvency resolution process related to insolvency and bankruptcy cases shall be considered ethical members and have the necessary qualifications for admission to the Insolvency and Bankruptcy Practice Resolution (CIL) follows updated with the latest legal degrees. The insolvency and organization of insolvency and liquidacy per Article 219 of the Indian Constitution shall continue to operate within the bankruptcy area for which it was not enacted. Other sectors as well may include Lincolving Reforms in terms of winding up proceedings in case of insolvency or liquidation proceedings fail Provide Complaints Committee to reconsider their decisions in cases involving insolvency professionals and courts.
S.3 What are the legal risks associated with starting a gambling business in India?
Response:
Under Section 637 of the Insolvency and Bankruptcy Code, 2016, in cases of insolvency and bankruptcy cases, all persons interested in a liquidy or insolvency resolution process related to insolvency and bankruptcy cases shall be considered ethical members and have the necessary qualifications for admission to the Insolvency and Bankruptcy Practice Resolution (CIL) follows updated with the latest legal degrees. The insolvency and organization of insolvency and liquidacy per Article 219 of the Indian Constitution shall continue to operate within the bankruptcy area for which it was not enacted. Other sectors as well may include Lincolving Reforms in terms of winding up proceedings in case of insolvency or liquidation proceedings fail Provide Complaints Committee to reconsider their decisions in cases involving insolvency professionals and courts.