Update README.md

README.md

@@ -1,75 +1,51 @@
 ---
 license: other
-base_model: hon9kon9ize/Qwen2.5-32B-cpt
+library_name: transformers
 tags:
 - llama-factory
 - full
 - generated_from_trainer
+base_model: hon9kon9ize/CantoneseLLM-v1.0-32B-cpt
 model-index:
-- name: Qwen2.5-32B-sft
+- name: CantoneseLLMChat-v1.0-32B
   results: []
 ---
 
-<!-- This model card has been generated automatically according to the information the Trainer had access to. You
-should probably proofread and complete it, then remove this comment. -->
 
-# Qwen2.5-32B-sft
+# CantoneseLLMChat-v1.0-32B
 
-This model is a fine-tuned version of [hon9kon9ize/Qwen2.5-32B-cpt](https://huggingface.co/hon9kon9ize/Qwen2.5-32B-cpt) on the sft_v1 dataset.
-It achieves the following results on the evaluation set:
-- Loss: 1.0515
+![front_image](cantonese_llm_v1.jpg)
 
-## Model description
-
-More information needed
-
-## Intended uses & limitations
-
-More information needed
-
-## Training and evaluation data
-
-More information needed
-
-## Training procedure
 
-### Training hyperparameters
+Cantonese LLM Chat v1.0 is the first generation Cantonese LLM from hon9kon9ize.
+Building upon the sucess of [v0.5 preview](https://huggingface.co/hon9kon9ize/CantoneseLLMChat-v0.5), the model excels in Hong Kong related specific knowledge and Cantonese conversation.
 
-The following hyperparameters were used during training:
-- learning_rate: 1e-05
-- train_batch_size: 2
-- eval_batch_size: 2
-- seed: 42
-- distributed_type: multi-GPU
-- num_devices: 16
-- gradient_accumulation_steps: 4
-- total_train_batch_size: 128
-- total_eval_batch_size: 32
-- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
-- lr_scheduler_type: cosine
-- lr_scheduler_warmup_ratio: 0.1
-- num_epochs: 3.0
-
-### Training results
-
-| Training Loss | Epoch  | Step | Validation Loss |
-|:-------------:|:------:|:----:|:---------------:|
-| 1.042         | 0.2676 | 100  | 1.0254          |
-| 0.9872        | 0.5351 | 200  | 1.0064          |
-| 1.008         | 0.8027 | 300  | 0.9934          |
-| 0.7473        | 1.0702 | 400  | 1.0106          |
-| 0.7788        | 1.3378 | 500  | 1.0046          |
-| 0.7246        | 1.6054 | 600  | 1.0002          |
-| 0.7525        | 1.8729 | 700  | 0.9971          |
-| 0.529         | 2.1405 | 800  | 1.0470          |
-| 0.5365        | 2.4080 | 900  | 1.0517          |
-| 0.5256        | 2.6756 | 1000 | 1.0514          |
-| 0.518         | 2.9431 | 1100 | 1.0516          |
-
-
-### Framework versions
-
-- Transformers 4.43.3
-- Pytorch 2.3.1+cu121
-- Datasets 2.20.0
-- Tokenizers 0.19.1
+## Model description
+Base model obtained via Continuous Pre-Training of [Qwen 2.5 32B](https://huggingface.co/Qwen/Qwen2.5-32B) with 600 millions publicaly available Hong Kong news articles and Cantonese websites. 
+Instructions fine-tuned model trained with a dataset consists of 75,000 instrutions pairs. 45,000 pairs were Cantonese insturctions generated by other LLMs and reviewed by humans. 
+
+The model trained with 1 Nvidia H100 80GB HBM3 GPU on [Genkai Supercomputer](https://www.cc.kyushu-u.ac.jp/scp/eng/system/Genkai/hardware/).
+
+## Basic Usage
+```
+import torch
+from transformers import AutoTokenizer, AutoModelForCausalLM
+model_id = "hon9kon9ize/CantoneseLLMChat-v1.0-32B"
+tokenizer = AutoTokenizer.from_pretrained(model_id)
+model = AutoModelForCausalLM.from_pretrained(
+    model_id,
+    torch_dtype=torch.bfloat16, 
+    device_map="auto", 
+)
+def chat(messages, temperature=0.9, max_new_tokens=200):
+    input_ids = tokenizer.apply_chat_template(conversation=messages, tokenize=True, add_generation_prompt=True, return_tensors='pt').to('cuda:0')
+    output_ids = model.generate(input_ids, max_new_tokens=max_new_tokens, temperature=temperature)
+    response = tokenizer.decode(output_ids[0][input_ids.shape[1]:], skip_special_tokens=False)
+    return response
+prompt = "邊個係香港特首？"
+messages = [
+    {"role": "system", "content": "you are a helpful assistant."},
+    {"role": "user", "content": prompt}
+]
+print(chat(messages)) # 香港特別行政區行政長官係李家超。<|im_end|>
+```