|
|
--- |
|
|
language: |
|
|
- ta |
|
|
license: apache-2.0 |
|
|
tags: |
|
|
- pretrained |
|
|
datasets: |
|
|
- Hemanth-thunder/tamil-madlad-400 |
|
|
pipeline_tag: text-generation |
|
|
inference: |
|
|
parameters: |
|
|
temperature: 0.7 |
|
|
--- |
|
|
# Model Card for Tamil-Mistral-7B-v0.1 |
|
|
|
|
|
The Tamil-Mistral-7B-v0.1 Large Language Model (LLM) is a pre-trained generative text model trained at the top of mistral base model 7 billion parameters. This is extends version of tokenization capability by increasing tamil tokens by 20k. |
|
|
Additionally, it was Pretrained on 1.19 million Tamil documents sourced from madlad-400 (Tamil) [MADLAD-400 (Multilingual Audited Dataset: Low-resource And Document-level)](https://arxiv.org/abs/2309.04662). |
|
|
|
|
|
pretraining time: 145 hours (GPU NVIDIA RTX A6000 48GB) |
|
|
## Mistral model details |
|
|
|
|
|
For full details of this model please read our [paper](https://arxiv.org/abs/2310.06825) and [release blog post](https://mistral.ai/news/announcing-mistral-7b/). |
|
|
|
|
|
## Model Architecture |
|
|
|
|
|
Mistral-7B-v0.1 is a transformer model, with the following architecture choices: |
|
|
- Grouped-Query Attention |
|
|
- Sliding-Window Attention |
|
|
- Byte-fallback BPE tokenizer |
|
|
|
|
|
|
|
|
#### Running the model on a GPU |
|
|
|
|
|
|
|
|
```python |
|
|
import torch |
|
|
from transformers import (AutoModelForCausalLM,AutoTokenizer,BitsAndBytesConfig,TextStreamer,pipeline) |
|
|
bnb_config = BitsAndBytesConfig(load_in_4bit=True,bnb_4bit_quant_type="nf4",bnb_4bit_compute_dtype="float16",bnb_4bit_use_double_quant=False,) |
|
|
model = AutoModelForCausalLM.from_pretrained("Hemanth-thunder/Tamil-Mistral-7B-v0.1",quantization_config=bnb_config,device_map={"": 0}) |
|
|
tokenizer = AutoTokenizer.from_pretrained("Hemanth-thunder/Tamil-Mistral-7B-v0.1",trust_remote_code=True) |
|
|
tokenizer.pad_token = tokenizer.eos_token |
|
|
tokenizer.padding_side = "right" |
|
|
streamer = TextStreamer(tokenizer) |
|
|
pipe = pipeline("text-generation" ,model=model, tokenizer=tokenizer ,do_sample=True, repetition_penalty=1.15,top_p=0.95,streamer=streamer) |
|
|
pipe("ஐபிஎல் தொடரில் மும்பை இந்தியன்ஸ் அணி ",max_length=50) |
|
|
``` |
|
|
|
|
|
```Generated text |
|
|
ஐபிஎல் தொடரில் மும்பை இந்தியன்ஸ் அணி ரிக்கி பாண்டிங்கை எதிர்த்து வருகிறது. இவ்விரு அணிகளுமே பலம் வாய்ந்தவை. ஆகவே, பரபரப்பான இந்த தொடரை பொறுத்தமட்டில் பிளே-ஆஃப் சுற்று வரை நுழைய வேண்டிய கட்டாயத்தில் இரு அணிகளும் உள்ளன. மும்பை இந்தியன்ஸ் அணி கடைசி லீக் போட்டியில் |
|
|
|
|
|
``` |
|
|
|
|
|
## Troubleshooting |
|
|
|
|
|
- If you see the following error: |
|
|
``` |
|
|
KeyError: 'mistral' |
|
|
``` |
|
|
- Or: |
|
|
``` |
|
|
NotImplementedError: Cannot copy out of meta tensor; no data! |
|
|
``` |
|
|
|
|
|
Ensure you are utilizing a stable version of Transformers, 4.34.0 or newer. |
|
|
|
|
|
## Notice |
|
|
|
|
|
Mistral 7B is a pretrained base model and therefore does not have any moderation mechanisms. |
|
|
|
|
|
## The Mistral AI Team |
|
|
|
|
|
Albert Jiang, Alexandre Sablayrolles, Arthur Mensch, Chris Bamford, Devendra Singh Chaplot, Diego de las Casas, Florian Bressand, Gianna Lengyel, Guillaume Lample, Lélio Renard Lavaud, Lucile Saulnier, Marie-Anne Lachaux, Pierre Stock, Teven Le Scao, Thibaut Lavril, Thomas Wang, Timothée Lacroix, William El Sayed. |
