Training in progress, step 50

config.json

@@ -1,5 +1,5 @@
 {
-  "_name_or_path": "/work/bmb/hf_models/openr1/DeepSeek-R1-Distill-Qwen-1.5B-GRPO-beta-0.04",
+  "_name_or_path": "/work/bmb/hf_models/Qwen/Qwen2.5-Math-1.5B-Instruct",
   "architectures": [
     "Qwen2ForCausalLM"
   ],

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5e7475ef2edb68e70e8d698526ec225e4c63e6e5311be8a7aac2806061c8983f
+oid sha256:79ed37c0e75a8f083d29ba3cf67cdca706b821bffb075d26c132e1f75f3a7207
 size 3554214752

tokenizer_config.json

@@ -200,12 +200,8 @@
   "eos_token": "<|im_end|>",
   "errors": "replace",
   "extra_special_tokens": {},
-  "max_length": null,
   "model_max_length": 131072,
-  "pad_to_multiple_of": null,
   "pad_token": "<|endoftext|>",
-  "pad_token_type_id": 0,
-  "padding_side": "left",
   "split_special_tokens": false,
   "tokenizer_class": "Qwen2Tokenizer",
   "unk_token": null

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:cfde4e89983697b6286a46e88ca7b2e4c2863ec5823bb968abb668e1b166b2ef
+oid sha256:6f642d0d845f7ab297fb4cfb3b81476d48a3417bc468255b997008ec5958d9f3
 size 8184