Upload Qwen2ForCausalLM

config.json

@@ -1,5 +1,5 @@
 {
-  "_name_or_path": "final_run_qwen/run-5/checkpoint-7656",
+  "_name_or_path": "final_run_qwen_0.5/run-0/checkpoint-8800",
   "architectures": [
     "Qwen2ForCausalLM"
   ],
@@ -7,21 +7,22 @@
   "bos_token_id": 151643,
   "eos_token_id": 151643,
   "hidden_act": "silu",
-  "hidden_size": 1536,
+  "hidden_size": 896,
   "initializer_range": 0.02,
-  "intermediate_size": 8960,
+  "intermediate_size": 4864,
   "max_position_embeddings": 131072,
-  "max_window_layers": 28,
+  "max_window_layers": 24,
   "model_type": "qwen2",
-  "num_attention_heads": 12,
-  "num_hidden_layers": 28,
+  "num_attention_heads": 14,
+  "num_hidden_layers": 24,
   "num_key_value_heads": 2,
   "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
   "rope_theta": 1000000.0,
   "sliding_window": null,
   "tie_word_embeddings": true,
   "torch_dtype": "float16",
-  "transformers_version": "4.44.2",
+  "transformers_version": "4.46.3",
   "use_cache": true,
   "use_sliding_window": false,
   "vocab_size": 151936

generation_config.json

@@ -2,5 +2,5 @@
   "bos_token_id": 151643,
   "eos_token_id": 151643,
   "max_new_tokens": 2048,
-  "transformers_version": "4.44.2"
+  "transformers_version": "4.46.3"
 }

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fa1b1f712e03ca426d522b3779d9b6a0028c9f2fe68fccc678b5d6d681605a3f
-size 3087466808
+oid sha256:720218e058f2c5a27af8881f0b653dab4b20134528ce16628db1e22d8f453bae
+size 988097536