Upload README.md with huggingface_hub

README.md

@@ -1,17 +1,19 @@
 ---
-license: apache-2.0
 library_name: peft
 base_model: meta-llama/Llama-2-7b-hf
 tags:
   - lora
-  - adapter
+  - peft
   - causal-lm
+  - adapter
+license: apache-2.0
 ---
 
-# Adapter Checkpoint — LoRA fine-tune on Llama-2-7b
+# Adapter Checkpoint — LoRA on Llama-2-7b
 
-A lightweight **LoRA** adapter trained on top of `meta-llama/Llama-2-7b-hf` for conversational tasks.  
-Load it with [🤗 PEFT](https://github.com/huggingface/peft) in a single line — no full model retraining required.
+This repository contains a **LoRA adapter checkpoint** fine-tuned on top of
+[`meta-llama/Llama-2-7b-hf`](https://huggingface.co/meta-llama/Llama-2-7b-hf)
+using [PEFT](https://github.com/huggingface/peft).
 
 ---
 
@@ -19,97 +21,78 @@ Load it with [🤗 PEFT](https://github.com/huggingface/peft) in a single line
 
 ```
 .
-├── adapter_config.json                   # PEFT / LoRA hyper-parameters
-├── adapter_model.bin                     # Trained adapter weights
-├── README.md
+├── adapter_config.json          # PEFT / LoRA hyper-parameters
+├── adapter_model.bin            # Trained adapter weights
+├── README.md                    # This file
 └── examples/
     └── chat/
         ├── zero_shot/
-        │   └── prompt.json               # Zero-shot chat prompt template
+        │   └── prompt.json      # Zero-shot chat prompt template
         └── few_shot/
-            └── prompt.json               # Few-shot chat prompt template
+            └── prompt.json      # Few-shot chat prompt template
 ```
 
 ---
 
-## Quick start
-
-```python
-from peft import PeftModel, PeftConfig
-from transformers import AutoModelForCausalLM, AutoTokenizer
+## Prompt templates
 
-repo_id = "dongbobo/adapter-checkpoint"
+Two ready-to-use prompt templates are included for chat inference:
 
-config    = PeftConfig.from_pretrained(repo_id)
-tokenizer = AutoTokenizer.from_pretrained(config.base_model_name_or_path)
-base      = AutoModelForCausalLM.from_pretrained(config.base_model_name_or_path)
-model     = PeftModel.from_pretrained(base, repo_id)
-model.eval()
-```
+| Strategy | Path | Description |
+|---|---|---|
+| Zero-shot | [`examples/chat/zero_shot/prompt.json`](examples/chat/zero_shot/prompt.json) | Single-turn; no demonstrations — the model relies on its instruction-following capability. |
+| Few-shot | [`examples/chat/few_shot/prompt.json`](examples/chat/few_shot/prompt.json) | Prepends three (user, assistant) demonstration turns before the live query. |
 
 ---
 
-## Prompt templates
-
-Two ready-to-use chat prompt templates are included.  
-Load either one and substitute `{{user_message}}` with your query.
-
-| Strategy | File |
-|---|---|
-| **Zero-shot** — no in-context examples | [`examples/chat/zero_shot/prompt.json`](examples/chat/zero_shot/prompt.json) |
-| **Few-shot** — three curated demonstrations | [`examples/chat/few_shot/prompt.json`](examples/chat/few_shot/prompt.json) |
-
-### Zero-shot usage
+## Quick start
 
 ```python
-import json, requests
+from peft import PeftModel, PeftConfig
+from transformers import AutoModelForCausalLM, AutoTokenizer
+import json, pathlib
 
-raw = requests.get(
-    "https://huggingface.co/dongbobo/adapter-checkpoint/resolve/main/"
-    "examples/chat/zero_shot/prompt.json"
-).json()
+# Load adapter config and base model
+config = PeftConfig.from_pretrained("dongbobo/adapter-checkpoint")
+base   = AutoModelForCausalLM.from_pretrained(config.base_model_name_or_path)
+model  = PeftModel.from_pretrained(base, "dongbobo/adapter-checkpoint")
+tok    = AutoTokenizer.from_pretrained(config.base_model_name_or_path)
 
-prompt = (
-    raw["template"]["system"] + "\n\n"
-    + raw["template"]["user"].replace("{{user_message}}", "Explain black holes briefly.")
+# Load a prompt template
+template = json.loads(
+    pathlib.Path("examples/chat/zero_shot/prompt.json").read_text()
 )
-```
-
-### Few-shot usage
 
-```python
-raw = requests.get(
-    "https://huggingface.co/dongbobo/adapter-checkpoint/resolve/main/"
-    "examples/chat/few_shot/prompt.json"
-).json()
-
-shots  = "\n".join(
-    f"User: {ex['user']}\nAssistant: {ex['assistant']}"
-    for ex in raw["template"]["examples"]
-)
-prompt = (
-    raw["template"]["system"] + "\n\n"
-    + shots + "\n\n"
-    + "User: " + "Explain black holes briefly." + "\nAssistant:"
+# Build prompt
+user_msg = "Explain the concept of attention in transformers."
+prompt   = (
+    f"<s>[INST] <<SYS>>\n{template['template']['system']}\n<</SYS>>\n\n"
+    f"{user_msg} [/INST]"
 )
+
+inputs  = tok(prompt, return_tensors="pt")
+outputs = model.generate(**inputs, max_new_tokens=256)
+print(tok.decode(outputs[0], skip_special_tokens=True))
 ```
 
 ---
 
-## Adapter configuration
+## Adapter hyper-parameters
 
 | Parameter | Value |
 |---|---|
-| PEFT type | `LORA` |
-| Rank (`r`) | 8 |
-| Alpha | 16 |
-| Dropout | 0.05 |
+| PEFT type | LORA |
+| Task type | CAUSAL\_LM |
+| Rank (`r`) | 16 |
+| LoRA alpha | 32 |
+| LoRA dropout | 0.05 |
 | Target modules | `q_proj`, `v_proj` |
-| Task type | `CAUSAL_LM` |
-| Base model | `meta-llama/Llama-2-7b-hf` |
+| Bias | none |
 
 ---
 
 ## License
 
-Released under the **Apache 2.0** licence. See [LICENSE](https://www.apache.org/licenses/LICENSE-2.0) for details.
+Released under the **Apache 2.0** license.  
+The base model (`meta-llama/Llama-2-7b-hf`) is subject to its own
+[Llama 2 Community License](https://huggingface.co/meta-llama/Llama-2-7b-hf/blob/main/LICENSE.txt).