Upload 3 files

README.md

@@ -1,3 +1,93 @@
----
-license: apache-2.0
----
+---
+base_model: llm-jp/llm-jp-3-13b
+library_name: peft
+---
+
+# 概要
+[llm-jp/llm-jp-3-13b](https://huggingface.co/llm-jp/llm-jp-3-13b) を [ichikara-instruction](https://liat-aip.sakura.ne.jp/wp/llm%e3%81%ae%e3%81%9f%e3%82%81%e3%81%ae%e6%97%a5%e6%9c%ac%e8%aa%9e%e3%82%a4%e3%83%b3%e3%82%b9%e3%83%88%e3%83%a9%e3%82%af%e3%82%b7%e3%83%a7%e3%83%b3%e3%83%87%e3%83%bc%e3%82%bf%e4%bd%9c%e6%88%90/llm%e3%81%ae%e3%81%9f%e3%82%81%e3%81%ae%e6%97%a5%e6%9c%ac%e8%aa%9e%e3%82%a4%e3%83%b3%e3%82%b9%e3%83%88%e3%83%a9%e3%82%af%e3%82%b7%e3%83%a7%e3%83%b3%e3%83%87%e3%83%bc%e3%82%bf-%e5%85%ac%e9%96%8b) でSFTしたモデル。SFTの際は、モデルパラメータに対し8bit量子化を行ったQLoRAを用いている。
+
+
+# 推論方法
+本モデルを用いて `elyza-tasks-100-TV_0.jsonl` に対して推論する方法を示す。
+
+## データ
+`elyza-tasks-100-TV_0.jsonl` を事前にダウンロードする。
+
+## サンプルコード
+```
+import json
+import re
+
+import peft
+import torch
+import transformers
+
+
+def load_jsonl(fname):
+    with open(fname, encoding="utf-8") as f:
+        data = []
+        for line in f:
+            _data = json.loads(line.strip())
+            data.append(_data)
+    return data
+
+
+# loading dataset
+dataset = load_jsonl("./elyza-tasks-100-TV_0.jsonl")
+
+
+# loading model
+bnb_config = transformers.BitsAndBytesConfig(load_in_8bit=True)
+
+model = transformers.AutoModelForCausalLM.from_pretrained(
+    pretrained_model_name_or_path="llm-jp/llm-jp-3-13b", device_map="auto", quantization_config=bnb_config
+)
+model = peft.PeftModel.from_pretrained(model, "orihihsoy/llm-jp-3-13b_qlora_8bit")
+
+tokenizer = transformers.AutoTokenizer.from_pretrained(
+    pretrained_model_name_or_path=="llm-jp/llm-jp-3-13b"
+)
+
+
+# evaluation
+PROMPT_TEMPLATE = """{instruction}
+
+### 指示:
+{input}
+
+### 回答:
+{output}"""
+
+results = []
+for data in dataset:
+    input = data["input"]
+    BOS_TOKEN = tokenizer.bos_token
+
+    prompt = BOS_TOKEN + PROMPT_TEMPLATE.format(
+        instruction="以下は、タスクを説明する指示です。要求を適切に満たす応答を書きなさい。", input=input, output="")
+
+    tokenized_input = tokenizer.encode(
+        prompt, add_special_tokens=False, return_tensors="pt").to(model.device)
+    attention_mask = torch.ones_like(tokenized_input)
+    with torch.no_grad():
+        outputs = model.generate(
+            tokenized_input,
+            attention_mask=attention_mask,
+            max_new_tokens=1024,
+            do_sample=True,
+            top_p=0.95,
+            temperature=0.7,
+            repetition_penalty=1.05,
+            pad_token_id=tokenizer.eos_token_id
+        )[0]
+    output = tokenizer.decode(
+        outputs[tokenized_input.size(1):], skip_special_tokens=True)
+
+    results.append({"task_id": data["task_id"],
+                    "input": input, "output": output})
+
+with open(f"gen.jsonl", 'w', encoding='utf-8') as f:
+    for result in results:
+        json.dump(result, f, ensure_ascii=False)
+        f.write('\n')
+```

adapter_config.json

@@ -0,0 +1,39 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "llm-jp/llm-jp-3-13b",
+  "bias": "none",
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 32,
+  "lora_bias": false,
+  "lora_dropout": 0.05,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 16,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "q_proj",
+    "gate_proj",
+    "up_proj",
+    "lm_head",
+    "o_proj",
+    "v_proj",
+    "down_proj",
+    "k_proj",
+    "embed_tokens"
+  ],
+  "task_type": "CAUSAL_LM",
+  "use_dora": false,
+  "use_rslora": false
+}

adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b0ad9ea0b37ec9ebc76630ffcb0c9ae804aa0972e78ba38f996a0bca2041410a
+size 2303306144