Trained challenger (loss=3.2392, step=3)

config.json

@@ -62,7 +62,7 @@
   "sliding_window": 512,
   "tie_word_embeddings": true,
   "top_k_experts": null,
-  "transformers_version": "5.5.4",
+  "transformers_version": "5.5.3",
   "use_bidirectional_attention": null,
   "use_cache": true,
   "use_double_wide_mlp": false,

generation_config.json

@@ -1,8 +1,8 @@
 {
   "_from_model_config": true,
   "bos_token_id": 2,
-  "eos_token_id": 106,
+  "eos_token_id": 1,
   "pad_token_id": 0,
-  "transformers_version": "5.5.4",
-  "use_cache": false
+  "transformers_version": "5.5.3",
+  "use_cache": true
 }

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3f969c4b7f1a3eb9c231c2752cceea47e0ce74c18a5314be75fb8e314356745c
-size 6576171052
+oid sha256:3fe2459b54f605fa54397a388d00a52bd6956b68dd2621224edd7407df568a99
+size 6575835380

tokenizer_config.json

@@ -6,7 +6,7 @@
   "eoi_token": "<end_of_image>",
   "eos_token": "<end_of_turn>",
   "image_token": "<image_soft_token>",
-  "is_local": true,
+  "is_local": false,
   "mask_token": "<mask>",
   "model_max_length": 32768,
   "model_specific_special_tokens": {