Model save

README.md

@@ -27,7 +27,7 @@ print(output["generated_text"])
 
 ## Training procedure
 
-[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="150" height="24"/>](https://wandb.ai/chloeli/huggingface/runs/bvttf6gb) 
+[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="150" height="24"/>](https://wandb.ai/chloeli/huggingface/runs/vj7aj0mi) 
 
 
 This model was trained with SFT.

adapter_config.json

@@ -23,13 +23,13 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "v_proj",
-    "o_proj",
     "gate_proj",
-    "q_proj",
-    "down_proj",
     "k_proj",
-    "up_proj"
+    "down_proj",
+    "up_proj",
+    "v_proj",
+    "o_proj",
+    "q_proj"
   ],
   "task_type": "CAUSAL_LM",
   "use_dora": false,

all_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 8.539757562822656e+16,
-    "train_loss": 0.07947162318229675,
-    "train_runtime": 16205.2206,
+    "train_loss": 0.0,
+    "train_runtime": 4.6206,
     "train_samples": 1000,
-    "train_samples_per_second": 0.062,
-    "train_steps_per_second": 0.008
+    "train_samples_per_second": 216.423,
+    "train_steps_per_second": 27.053
 }

config.json

@@ -0,0 +1,45 @@
+{
+  "_attn_implementation_autoset": true,
+  "_name_or_path": "Qwen/Qwen2.5-7B-Instruct",
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "hidden_act": "silu",
+  "hidden_size": 3584,
+  "initializer_range": 0.02,
+  "intermediate_size": 18944,
+  "max_position_embeddings": 32768,
+  "max_window_layers": 28,
+  "model_type": "qwen2",
+  "num_attention_heads": 28,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 4,
+  "quantization_config": {
+    "_load_in_4bit": true,
+    "_load_in_8bit": false,
+    "bnb_4bit_compute_dtype": "bfloat16",
+    "bnb_4bit_quant_storage": "uint8",
+    "bnb_4bit_quant_type": "nf4",
+    "bnb_4bit_use_double_quant": false,
+    "llm_int8_enable_fp32_cpu_offload": false,
+    "llm_int8_has_fp16_weight": false,
+    "llm_int8_skip_modules": null,
+    "llm_int8_threshold": 6.0,
+    "load_in_4bit": true,
+    "load_in_8bit": false,
+    "quant_method": "bitsandbytes"
+  },
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": 131072,
+  "tie_word_embeddings": false,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.49.0",
+  "use_cache": true,
+  "use_sliding_window": false,
+  "vocab_size": 152064
+}

train_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 8.539757562822656e+16,
-    "train_loss": 0.07947162318229675,
-    "train_runtime": 16205.2206,
+    "train_loss": 0.0,
+    "train_runtime": 4.6206,
     "train_samples": 1000,
-    "train_samples_per_second": 0.062,
-    "train_steps_per_second": 0.008
+    "train_samples_per_second": 216.423,
+    "train_steps_per_second": 27.053
 }

trainer_state.json

@@ -445,10 +445,10 @@
       "epoch": 1.0,
       "step": 125,
       "total_flos": 8.539757562822656e+16,
-      "train_loss": 0.07947162318229675,
-      "train_runtime": 16205.2206,
-      "train_samples_per_second": 0.062,
-      "train_steps_per_second": 0.008
+      "train_loss": 0.0,
+      "train_runtime": 4.6206,
+      "train_samples_per_second": 216.423,
+      "train_steps_per_second": 27.053
     }
   ],
   "logging_steps": 5,

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:567a9223e05e93aa37cca09150f45057a19d743d6b25ce3b7a100914805606db
+oid sha256:d02e82def1420c15f76ea6371f615868ed49d2bd98c83c835834baaa4b2e9076
 size 6456