Update README.md

README.md

@@ -1,22 +1,109 @@
 ---
 language:
 - en
-license: apache-2.0
+license: cc-by-nc-4.0
 tags:
 - text-generation-inference
 - transformers
 - unsloth
 - gemma
 - trl
+- finlang
+- qlora
 base_model: unsloth/gemma-7b-bnb-4bit
 ---
 
 # Uploaded  model
 
 - **Developed by:** anamikac2708
-- **License:** apache-2.0
+- **License:** cc-by-nc-4.0
 - **Finetuned from model :** unsloth/gemma-7b-bnb-4bit
 
-This gemma model was trained 2x faster with [Unsloth](https://github.com/unslothai/unsloth) and Huggingface's TRL library.
+This gemma model was trained 2x faster with [Unsloth](https://github.com/unslothai/unsloth) and Huggingface's TRL library using open-sourced finance dataset https://huggingface.co/datasets/FinLang/investopedia-instruction-tuning-dataset developed for finance application by FinLang Team
+
+This project is for research purposes only. Third-party datasets may be subject to additional terms and conditions under their associated licenses.
+
+## How to Get Started with the Model
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+```
+import torch
+from unsloth import FastLanguageModel
+from transformers import AutoTokenizer, pipeline
+model_id='FinLang/investopedia_chat_model'
+max_seq_length=2048
+model, tokenizer = FastLanguageModel.from_pretrained(
+        model_name = "anamikac2708/Gemma-7b-finetuned-investopedia-Merged-FP16", # YOUR MODEL YOU USED FOR TRAINING
+        max_seq_length = max_seq_length,
+        dtype = torch.bfloat16,
+        #load_in_4bit = True,  # IF YOU WANT TO LOAD WITH BITSANDBYTES INT4
+    )
+tokenizer = AutoTokenizer.from_pretrained(model_id)
+pipe = pipeline("text-generation", model=model, tokenizer=tokenizer)
+example = [{'content': 'You are a financial expert and you can answer any questions related to finance. You will be given a context and a question. Understand the given context and\n        try to answer. Users will ask you questions in English and you will generate answer based on the provided CONTEXT.\n        CONTEXT:\n        D. in Forced Migration from the University of the Witwatersrand (Wits) in Johannesburg, South Africa; A postgraduate diploma in Folklore & Cultural Studies at Indira Gandhi National Open University (IGNOU) in New Delhi, India; A Masters of International Affairs at Columbia University; A BA from Barnard College at Columbia University\n', 'role': 'system'}, {'content': ' In which universities did the individual obtain their academic qualifications?\n', 'role': 'user'}, {'content': ' University of the Witwatersrand (Wits) in Johannesburg, South Africa; Indira Gandhi National Open University (IGNOU) in New Delhi, India; Columbia University; Barnard College at Columbia University.', 'role': 'assistant'}]
+prompt = pipe.tokenizer.apply_chat_template(example[:2], tokenize=False, add_generation_prompt=True)
+outputs = pipe(prompt, max_new_tokens=256, do_sample=True, temperature=0.1, top_k=50, top_p=0.1, eos_token_id=pipe.tokenizer.eos_token_id, pad_token_id=pipe.tokenizer.pad_token_id)
+print(f"Query:\n{example[1]['content']}")
+print(f"Context:\n{example[0]['content']}")
+print(f"Original Answer:\n{example[2]['content']}")
+print(f"Generated Answer:\n{outputs[0]['generated_text'][len(prompt):].strip()}")
+```
+
+## Training Details
+```
+Peft Config :
+
+{
+ 'Technqiue' : 'QLORA',
+ 'rank': 256,
+ 'target_modules' : ["q_proj", "k_proj", "v_proj", "o_proj","gate_proj", "up_proj", "down_proj",],
+ 'lora_alpha' : 128,
+ 'lora_dropout' : 0, 
+ 'bias': "none",    
+}
+    
+Hyperparameters:
+
+{
+    "epochs": 3,
+    "evaluation_strategy": "epoch",
+    "gradient_checkpointing": True,
+    "max_grad_norm" : 0.3,
+    "optimizer" : "adamw_torch_fused",
+    "learning_rate" : 2e-5,
+    "lr_scheduler_type": "constant",
+    "warmup_ratio" : 0.03,
+    "per_device_train_batch_size" : 4,  
+    "per_device_eval_batch_size" : 4,
+    "gradient_accumulation_steps" : 4
+}
+```
+
+## Model was trained on 1xA100 80GB, below loss and memory consmuption details:
+{'eval_loss': 1.0056027173995972, 'eval_runtime': 299.1871, 'eval_samples_per_second': 2.276, 'eval_steps_per_second': 0.572, 'epoch': 3.0}
+{'train_runtime': 23623.219, 'train_samples_per_second': 0.683, 'train_steps_per_second': 0.043, 'train_loss': 0.8518931362777948, 'epoch': 3.0}
+Total training time 23623.95972943306
+23623.219 seconds used for training.
+393.72 minutes used for training.
+Peak reserved memory = 63.17 GB.
+Peak reserved memory for training = 54.553 GB.
+Peak reserved memory % of max memory = 79.809 %.
+Peak reserved memory for training % of max memory = 68.923 %.
+
+QLORA paper link - https://arxiv.org/abs/2305.14314
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+We evaluated the model on test set (sample 1k) https://huggingface.co/datasets/FinLang/investopedia-instruction-tuning-dataset. Evaluation was done using Proprietary LLMs as jury on four criteria Correctness, Faithfullness, Clarity, Completeness on scale of 1-5 (1 being worst & 5 being best) inspired  by the paper Replacing Judges with Juries https://arxiv.org/abs/2404.18796. Model got an average score of 3.84.
+Average inference speed of the model is 9.7 secs. Human Evaluation is in progress to see the percentage of alignment between human and LLM.
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+This model is a quick demonstration that the base model can be easily fine-tuned to achieve compelling performance. It does not have any moderation mechanisms. We're looking into ways to make the model finely respect guardrails, allowing for deployment in environments requiring moderated outputs.
+
+## License
+
+Since non-commercial datasets are used for fine-tuning, we release this model as cc-by-nc-4.0.
 
-[<img src="https://raw.githubusercontent.com/unslothai/unsloth/main/images/unsloth%20made%20with%20love.png" width="200"/>](https://github.com/unslothai/unsloth)