add awq model

config.json

@@ -0,0 +1,38 @@
+{
+  "_name_or_path": "/home/ubuntu/.cache/huggingface/hub/models--Qwen--CodeQwen1.5-7B/snapshots/3e2a9bf83397e9f9e8756df1a1ef10a8e97642de",
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 2,
+  "eos_token_id": 2,
+  "hidden_act": "silu",
+  "hidden_size": 4096,
+  "initializer_range": 0.02,
+  "intermediate_size": 13440,
+  "max_position_embeddings": 65536,
+  "max_window_layers": 28,
+  "model_type": "qwen2",
+  "num_attention_heads": 32,
+  "num_hidden_layers": 32,
+  "num_key_value_heads": 4,
+  "quantization_config": {
+    "bits": 4,
+    "group_size": 128,
+    "modules_to_not_convert": null,
+    "quant_method": "awq",
+    "version": "gemm",
+    "zero_point": true
+  },
+  "rms_norm_eps": 1e-05,
+  "rope_theta": 1000000,
+  "rotary_emb_base": 1000000,
+  "seq_length": 65536,
+  "sliding_window": 65536,
+  "tie_word_embeddings": false,
+  "torch_dtype": "float16",
+  "transformers_version": "4.38.2",
+  "use_cache": false,
+  "use_sliding_window": false,
+  "vocab_size": 92416
+}

generation_config.json

@@ -0,0 +1,11 @@
+{
+  "bos_token_id": 2,
+  "do_sample": true,
+  "eos_token_id": [
+    4,
+    2
+  ],
+  "pad_token_id": 92298,
+  "top_p": 0.95,
+  "transformers_version": "4.38.2"
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a5bc44aa9e116875f5d3d5b4b7cd04001fff6f9da9908cc3b9edd6e5eec5c524
+size 4888301576

quant_config.json

@@ -0,0 +1,6 @@
+{
+  "zero_point": true,
+  "q_group_size": 128,
+  "w_bit": 4,
+  "version": "GEMM"
+}

special_tokens_map.json

@@ -0,0 +1,38 @@
+{
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<fim_prefix>",
+    "<fim_middle>",
+    "<fim_suffix>",
+    "<fim_pad>"
+  ],
+  "bos_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<fim_pad>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "unk_token": {
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}