init

config.json

@@ -0,0 +1,53 @@
+{
+  "_name_or_path": "SushiTokyo/ELYZA-japanese-Llama-2-13b-fast-instruct-4bit-quantized",
+  "architectures": [
+    "LlamaForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "hidden_act": "silu",
+  "hidden_size": 5120,
+  "initializer_range": 0.02,
+  "intermediate_size": 13824,
+  "max_position_embeddings": 4096,
+  "model_type": "llama",
+  "num_attention_heads": 40,
+  "num_hidden_layers": 40,
+  "num_key_value_heads": 40,
+  "pad_token_id": 2,
+  "pretraining_tp": 1,
+  "quantization_config": {
+    "batch_size": 1,
+    "bits": 4,
+    "block_name_to_quantize": null,
+    "cache_block_outputs": true,
+    "damp_percent": 0.1,
+    "dataset": "c4",
+    "desc_act": false,
+    "exllama_config": {
+      "version": 1
+    },
+    "group_size": 64,
+    "max_input_length": null,
+    "model_seqlen": null,
+    "module_name_preceding_first_block": null,
+    "modules_in_block_to_quantize": null,
+    "pad_token_id": null,
+    "quant_method": "gptq",
+    "sym": true,
+    "tokenizer": null,
+    "true_sequential": true,
+    "use_cuda_fp16": false,
+    "use_exllama": true
+  },
+  "rms_norm_eps": 1e-05,
+  "rope_scaling": null,
+  "rope_theta": 10000.0,
+  "tie_word_embeddings": false,
+  "torch_dtype": "float16",
+  "transformers_version": "4.40.1",
+  "use_cache": true,
+  "vocab_size": 44581
+}

generation_config.json

@@ -0,0 +1,7 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "pad_token_id": 2,
+  "transformers_version": "4.40.1"
+}

model-00001-of-00002.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ee4a25030bfca8e7ed4bea9747664274082278e76308e2f3b55ac37199df80c1
+size 4966876440

model-00002-of-00002.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6c04e72e0fb82031456d9aa6efe99c1ba2bde9e706af028a89b80c00d1e4ccbe
+size 2798025296

quantize.py

@@ -0,0 +1,8 @@
+from transformers import AutoModelForCausalLM, AutoTokenizer, GPTQConfig
+model_id = "elyza/ELYZA-japanese-Llama-2-13b-fast-instruct"
+tokenizer = AutoTokenizer.from_pretrained(model_id)
+gptq_config = GPTQConfig(bits=4, dataset = "c4", tokenizer=tokenizer, group_size=64)
+model = AutoModelForCausalLM.from_pretrained(model_id, device_map="cuda", quantization_config=gptq_config)
+
+tokenizer.save_pretrained("./quantized")
+model.save_pretrained("./quantized")

readme.md

@@ -0,0 +1,12 @@
+## About
+ELYZA-japanese-Llama-2-13b-fast-instruct を4bitで量子化したモデルです。  
+ref: https://huggingface.co/elyza/ELYZA-japanese-Llama-2-13b-fast-instruct/blob/main/README.md  
+
+オリジナルはRTX4090でも扱いが難しかったですが、10秒ほどで出力できます。  
+精度は未計測。
+
+量子化の詳細はconfig.jsonを参照してください。  
+コードはquantize.pyです。  
+
+## LICENSE
+LICENSE: LLAMA2  

special_tokens_map.json

@@ -0,0 +1,35 @@
+{
+  "additional_special_tokens": [
+    "<unk>",
+    "<s>",
+    "</s>"
+  ],
+  "bos_token": {
+    "content": "<s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "unk_token": {
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

tokenizer.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:21715bf9f86ad0ba1236012f5aadee93239306fe0b87a448532e3299865fa36f
+size 705214

tokenizer_config.json

@@ -0,0 +1,47 @@
+{
+  "add_bos_token": true,
+  "add_eos_token": false,
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "<s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "additional_special_tokens": [
+    "<unk>",
+    "<s>",
+    "</s>"
+  ],
+  "bos_token": "<s>",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "</s>",
+  "legacy": true,
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": "</s>",
+  "padding_side": "right",
+  "sp_model_kwargs": {},
+  "spaces_between_special_tokens": false,
+  "tokenizer_class": "LlamaTokenizer",
+  "unk_token": "<unk>",
+  "use_default_system_prompt": false
+}