Trained with Unsloth

config.json

@@ -1,5 +1,5 @@
 {
-  "_name_or_path": "Qwen/Qwen2-0.5B-Instruct",
+  "_name_or_path": "Qwen/Qwen2-1.5B-Instruct",
   "architectures": [
     "Qwen2ForCausalLM"
   ],
@@ -7,14 +7,14 @@
   "bos_token_id": 151643,
   "eos_token_id": 151645,
   "hidden_act": "silu",
-  "hidden_size": 896,
+  "hidden_size": 1536,
   "initializer_range": 0.02,
-  "intermediate_size": 4864,
+  "intermediate_size": 8960,
   "max_position_embeddings": 32768,
-  "max_window_layers": 24,
+  "max_window_layers": 28,
   "model_type": "qwen2",
-  "num_attention_heads": 14,
-  "num_hidden_layers": 24,
+  "num_attention_heads": 12,
+  "num_hidden_layers": 28,
   "num_key_value_heads": 2,
   "rms_norm_eps": 1e-06,
   "rope_theta": 1000000.0,

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b89fc6bea4df00dbaaa26b0a32151c9d09772faf8e26cc22d375dcc7025fe7bd
-size 988034784
+oid sha256:418d4b9f32f85f8558b13820b4e485bba2adef2b62a9fe3e17d99cbd82f17d14
+size 3087343248