End of training

README.md

@@ -1,5 +1,5 @@
 ---
-base_model: Gensyn/Qwen2.5-0.5B-Instruct
+base_model: unsloth/Qwen2.5-0.5B-Instruct
 library_name: transformers
 model_name: Qwen2.5-0.5B-Instruct-Gensyn-Swarm-sly_nimble_lemur
 tags:
@@ -14,7 +14,7 @@ licence: license
 
 # Model Card for Qwen2.5-0.5B-Instruct-Gensyn-Swarm-sly_nimble_lemur
 
-This model is a fine-tuned version of [Gensyn/Qwen2.5-0.5B-Instruct](https://huggingface.co/Gensyn/Qwen2.5-0.5B-Instruct).
+This model is a fine-tuned version of [unsloth/Qwen2.5-0.5B-Instruct](https://huggingface.co/unsloth/Qwen2.5-0.5B-Instruct).
 It has been trained using [TRL](https://github.com/huggingface/trl).
 
 ## Quick start

all_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 6.447264695452759e-06,
-    "train_runtime": 10252.1281,
-    "train_samples": 17,
-    "train_samples_per_second": 0.016,
-    "train_steps_per_second": 0.001
+    "train_loss": 1.5658140182495117e-05,
+    "train_runtime": 2348.1696,
+    "train_samples": 4,
+    "train_samples_per_second": 0.009,
+    "train_steps_per_second": 0.002
 }

config.json

@@ -3,7 +3,6 @@
     "Qwen2ForCausalLM"
   ],
   "attention_dropout": 0.0,
-  "bos_token_id": 151643,
   "eos_token_id": 151645,
   "hidden_act": "silu",
   "hidden_size": 896,
@@ -15,13 +14,15 @@
   "num_attention_heads": 14,
   "num_hidden_layers": 24,
   "num_key_value_heads": 2,
+  "pad_token_id": 151654,
   "rms_norm_eps": 1e-06,
   "rope_scaling": null,
   "rope_theta": 1000000.0,
-  "sliding_window": 32768,
+  "sliding_window": null,
   "tie_word_embeddings": true,
   "torch_dtype": "float32",
   "transformers_version": "4.51.2",
+  "unsloth_fixed": true,
   "use_cache": true,
   "use_sliding_window": false,
   "vocab_size": 151936

generation_config.json

@@ -5,7 +5,8 @@
     151645,
     151643
   ],
-  "pad_token_id": 151643,
+  "max_length": 32768,
+  "pad_token_id": 151654,
   "repetition_penalty": 1.1,
   "temperature": 0.7,
   "top_k": 20,

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a1d1bfde27d5e6be495333e581102a6579a06b7e6a052a1a55513f1ec93bcc2f
+oid sha256:e3dd3620f46a1849ce2f350def4f07c571ec2c8b00086e326e6ec8a180f53071
 size 1976163472

special_tokens_map.json

@@ -22,7 +22,7 @@
     "single_word": false
   },
   "pad_token": {
-    "content": "<|endoftext|>",
+    "content": "<|vision_pad|>",
     "lstrip": false,
     "normalized": false,
     "rstrip": false,

tokenizer.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5eee858c5123a4279c3e1f7b81247343f356ac767940b2692a928ad929543214
-size 11422063
+oid sha256:64e71213db910f5cafa86d35091f37393dcc344b1bbc34091d1b3eed4cca01d5
+size 11422064

tokenizer_config.json

@@ -200,8 +200,9 @@
   "eos_token": "<|im_end|>",
   "errors": "replace",
   "extra_special_tokens": {},
-  "model_max_length": 131072,
-  "pad_token": "<|endoftext|>",
+  "model_max_length": 32768,
+  "pad_token": "<|vision_pad|>",
+  "padding_side": "left",
   "split_special_tokens": false,
   "tokenizer_class": "Qwen2Tokenizer",
   "unk_token": null

train_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 6.447264695452759e-06,
-    "train_runtime": 10252.1281,
-    "train_samples": 17,
-    "train_samples_per_second": 0.016,
-    "train_steps_per_second": 0.001
+    "train_loss": 1.5658140182495117e-05,
+    "train_runtime": 2348.1696,
+    "train_samples": 4,
+    "train_samples_per_second": 0.009,
+    "train_steps_per_second": 0.002
 }

trainer_state.json

@@ -2,123 +2,78 @@
   "best_global_step": null,
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 3.4705882352941178,
+  "epoch": 2.5,
   "eval_steps": 500,
-  "global_step": 10,
+  "global_step": 5,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "completion_length": 129.0625,
-      "epoch": 0.9411764705882353,
-      "grad_norm": 5.772158622741699,
+      "completion_length": 126.0,
+      "epoch": 1.0,
+      "grad_norm": 17.677051544189453,
       "kl": 0.0,
       "learning_rate": 5e-07,
       "loss": -0.0,
-      "reward": 0.591631580144167,
-      "reward_std": 0.2950996853032848,
+      "reward": 0.09396408637985587,
+      "reward_std": 0.02489092040923424,
       "rewards/concensus_correctness_reward_func": 0.0,
       "rewards/consensus_reward_func": 0.0,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0625,
-      "rewards/question_recreation_reward_func": 0.500725319609046,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.07896408578380942,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.02840625075623393,
+      "rewards/xmlcount_reward_func": 0.014999999664723873,
       "step": 2
     },
     {
-      "completion_length": 129.88888888888889,
-      "epoch": 1.4705882352941178,
-      "grad_norm": 4.538054943084717,
-      "kl": 0.003908961649156279,
-      "learning_rate": 4.415111107797445e-07,
+      "completion_length": 93.5,
+      "epoch": 2.0,
+      "grad_norm": 22.663196563720703,
+      "kl": 0.02060387669780539,
+      "learning_rate": 2.5e-07,
       "loss": 0.0,
-      "reward": 0.5140680852863524,
-      "reward_std": 0.10897246843928264,
+      "reward": 0.058309531887061894,
+      "reward_std": 0.02309032447374193,
       "rewards/concensus_correctness_reward_func": 0.0,
       "rewards/consensus_reward_func": 0.0,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.49312362840606105,
+      "rewards/question_recreation_reward_func": 0.08130953146610409,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.020944445704420406,
+      "rewards/xmlcount_reward_func": -0.023000000044703484,
       "step": 4
     },
     {
-      "completion_length": 138.77777777777777,
-      "epoch": 2.0,
-      "grad_norm": 5.477197647094727,
-      "kl": 0.008041509905726545,
-      "learning_rate": 2.934120444167326e-07,
-      "loss": 0.0,
-      "reward": 0.7325452516476313,
-      "reward_std": 0.1888044167103039,
-      "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.0,
-      "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.2222222222222222,
-      "rewards/question_recreation_reward_func": 0.5095452732510037,
-      "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.0007777793022493521,
-      "step": 6
-    },
-    {
-      "completion_length": 154.5,
-      "epoch": 2.9411764705882355,
-      "grad_norm": 4.317551136016846,
-      "kl": 0.016037766326917335,
-      "learning_rate": 1.2500000000000005e-07,
-      "loss": 0.0,
-      "reward": 0.5921503761783242,
-      "reward_std": 0.3880798064637929,
-      "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.0,
-      "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.125,
-      "rewards/question_recreation_reward_func": 0.5306503740139306,
-      "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": -0.0635000029578805,
-      "step": 8
-    },
-    {
-      "completion_length": 145.83333333333334,
-      "epoch": 3.4705882352941178,
-      "grad_norm": 5.180707931518555,
-      "kl": 0.016869528788245387,
-      "learning_rate": 1.507684480352292e-08,
-      "loss": 0.0,
-      "reward": 0.6629479693041908,
-      "reward_std": 0.28609387824932736,
+      "completion_length": 78.5,
+      "epoch": 2.5,
+      "kl": 0.03730800951598212,
+      "reward": 0.09967234451323748,
+      "reward_std": 0.01814919151365757,
       "rewards/concensus_correctness_reward_func": 0.0,
       "rewards/consensus_reward_func": 0.0,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.567670188844204,
+      "rewards/question_recreation_reward_func": 0.09967234451323748,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.09527777466509077,
-      "step": 10
-    },
-    {
-      "epoch": 3.4705882352941178,
-      "step": 10,
+      "rewards/xmlcount_reward_func": 0.0,
+      "step": 5,
       "total_flos": 0.0,
-      "train_loss": 6.447264695452759e-06,
-      "train_runtime": 10252.1281,
-      "train_samples_per_second": 0.016,
-      "train_steps_per_second": 0.001
+      "train_loss": 1.5658140182495117e-05,
+      "train_runtime": 2348.1696,
+      "train_samples_per_second": 0.009,
+      "train_steps_per_second": 0.002
     }
   ],
   "logging_steps": 2,
-  "max_steps": 10,
+  "max_steps": 5,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 5,
-  "save_steps": 25,
+  "num_train_epochs": 3,
+  "save_steps": 10,
   "stateful_callbacks": {
     "TrainerControl": {
       "args": {

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c1ae537136e5e8372083a14299b6eb3734101d7a7a4ec98e326d5b435ef1cb47
+oid sha256:32f80c30de5400cd309ee7238c1eeac7b9a967cb678460006d0b306cb7d16ec2
 size 5944