Update README.md

README.md

@@ -1,6 +1,6 @@
 ---
 language:
-- en
+- bn
 license: apache-2.0
 tags:
 - text-generation-inference
@@ -9,14 +9,110 @@ tags:
 - llama
 - trl
 base_model: unsloth/llama-3-8b-bnb-4bit
+datasets:
+- iamshnoo/alpaca-cleaned-bengali
+pipeline_tag: text-generation
 ---
 
-# Uploaded  model
+
+# LLama-3 Bangla
+
+<div align="center">
+    <img src="https://cdn-uploads.huggingface.co/production/uploads/65ca6f0098a46a56261ac3ac/O1ATwhQt_9j59CSIylrVS.png" width="300"/>
+
+</div>
 
 - **Developed by:** KillerShoaib
 - **License:** apache-2.0
 - **Finetuned from model :** unsloth/llama-3-8b-bnb-4bit
+- **Dataset used for finetuning :** iamshnoo/alpaca-cleaned-bengali
+
+
+
+# Model Details
+
+Llama 3 8 billion model was finetuned using **unsloth** package and in **4bit quantization** on a **cleaned Bangla alpaca** dataset. This is not the entire model but only **LoRA adapters**. The model is finetuned for **2 epoch** on a single T4 GPU.
+
+
+
+# Run The Model
+
+## FastLanguageModel from unsloth for 2x faster inference
+
+```python
+
+from unsloth import FastLanguageModel
+model, tokenizer = FastLanguageModel.from_pretrained(
+    model_name = "KillerShoaib/llama-3-8b-bangla-4bit",
+    max_seq_length = 2048,
+    dtype = None,
+    load_in_4bit = True,
+)
+FastLanguageModel.for_inference(model)
+
+# alpaca_prompt for the model
+alpaca_prompt = """Below is an instruction in bangla that describes a task, paired with an input also in bangla that provides further context. Write a response in bangla that appropriately completes the request.
+
+### Instruction:
+{}
+
+### Input:
+{}
+
+### Response:
+{}"""
+
+# input with instruction and input
+inputs = tokenizer(
+[
+    alpaca_prompt.format(
+        "সুস্থ থাকার তিনটি উপায় বলুন", # instruction
+        "", # input
+        "", # output - leave this blank for generation!
+    )
+], return_tensors = "pt").to("cuda")
+
+# generating the output and decoding it
+outputs = model.generate(**inputs, max_new_tokens = 2048, use_cache = True)
+tokenizer.batch_decode(outputs)
+```
+
+## AutoModelForPeftCausalLM from Hugginface
+
+```python
+from peft import AutoPeftModelForCausalLM
+from transformers import AutoTokenizer
+load_in_4bit = True
+model = AutoPeftModelForCausalLM.from_pretrained(
+    "KillerShoaib/llama3-8b-4bit-bangla",
+    load_in_4bit = True,
+)
+tokenizer = AutoTokenizer.from_pretrained("KillerShoaib/llama3-8b-4bit-bangla")
+
+alpaca_prompt = """Below is an instruction in bangla that describes a task, paired with an input also in bangla that provides further context. Write a response in bangla that appropriately completes the request.
+
+### Instruction:
+{}
+
+### Input:
+{}
+
+### Response:
+{}"""
+
+inputs = tokenizer(
+[
+    alpaca_prompt.format(
+        "সুস্থ থাকার তিনটি উপায় বলুন", # instruction
+        "", # input
+        "", # output - leave this blank for generation!
+    )
+], return_tensors = "pt").to("cuda")
+
+outputs = model.generate(**inputs, max_new_tokens = 1024, use_cache = True)
+tokenizer.batch_decode(outputs)
+```
+
 
-This llama model was trained 2x faster with [Unsloth](https://github.com/unslothai/unsloth) and Huggingface's TRL library.
+**AutoModelForPeftCausalLM can be hopelessly slow, since `4bit` model downloading is not supported. Use this only if you don't have unsloth installed**
 
-[<img src="https://raw.githubusercontent.com/unslothai/unsloth/main/images/unsloth%20made%20with%20love.png" width="200"/>](https://github.com/unslothai/unsloth)