End of training

adapter_config.json

@@ -23,22 +23,22 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "audio_enc_to_dec_proj",
-    "embed_tokens.1",
-    "enc_to_dec_proj",
-    "q_proj",
-    "lm_heads.1",
-    "v_proj",
+    "embed_tokens.2",
     "out_proj",
+    "embed_tokens.0",
+    "lm_heads.1",
+    "q_proj",
+    "fc2",
+    "lm_heads.2",
     "lm_heads.0",
-    "k_proj",
     "lm_heads.3",
-    "fc2",
-    "fc1",
+    "enc_to_dec_proj",
+    "embed_tokens.1",
+    "audio_enc_to_dec_proj",
+    "v_proj",
     "embed_tokens.3",
-    "embed_tokens.0",
-    "embed_tokens.2",
-    "lm_heads.2"
+    "k_proj",
+    "fc1"
   ],
   "task_type": null,
   "use_dora": false,

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c3612eb015af73cbade088639234ac52de0079fef04916f606aaabbb57062e82
+oid sha256:d8b947692009e5564f54e1a1c8064d476f99c0931c814df934f82b46cbdd7b52
 size 87103456

trainer_state.json

@@ -3,65 +3,107 @@
   "best_model_checkpoint": null,
   "epoch": 3.84,
   "eval_steps": 500,
-  "global_step": 12,
+  "global_step": 24,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0525139570236206,
+      "learning_rate": 0.00018333333333333334,
+      "loss": 9.5169,
+      "step": 2
+    },
     {
       "epoch": 0.64,
-      "grad_norm": 1.2096633911132812,
+      "grad_norm": 1.5870144367218018,
       "learning_rate": 0.0001666666666666667,
-      "loss": 9.4081,
-      "step": 2
+      "loss": 9.1514,
+      "step": 4
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7580112218856812,
+      "learning_rate": 0.00015000000000000001,
+      "loss": 8.7557,
+      "step": 6
     },
     {
       "epoch": 1.28,
-      "grad_norm": 2.210158109664917,
+      "grad_norm": 2.0391345024108887,
       "learning_rate": 0.00013333333333333334,
-      "loss": 8.9596,
-      "step": 4
+      "loss": 8.2163,
+      "step": 8
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 2.67323899269104,
+      "learning_rate": 0.00011666666666666668,
+      "loss": 7.9266,
+      "step": 10
     },
     {
       "epoch": 1.92,
-      "grad_norm": 1.7887189388275146,
+      "grad_norm": 1.5127941370010376,
       "learning_rate": 0.0001,
-      "loss": 8.5658,
-      "step": 6
+      "loss": 7.6902,
+      "step": 12
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.308194875717163,
+      "learning_rate": 8.333333333333334e-05,
+      "loss": 7.5201,
+      "step": 14
     },
     {
       "epoch": 2.56,
-      "grad_norm": 1.780813455581665,
+      "grad_norm": 0.9491788148880005,
       "learning_rate": 6.666666666666667e-05,
-      "loss": 8.1075,
-      "step": 8
+      "loss": 7.3685,
+      "step": 16
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.15030837059021,
+      "learning_rate": 5e-05,
+      "loss": 7.3428,
+      "step": 18
     },
     {
       "epoch": 3.2,
-      "grad_norm": 1.9974435567855835,
+      "grad_norm": 0.9033100008964539,
       "learning_rate": 3.3333333333333335e-05,
-      "loss": 7.9692,
-      "step": 10
+      "loss": 7.2229,
+      "step": 20
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 1.4382227659225464,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 7.3239,
+      "step": 22
     },
     {
       "epoch": 3.84,
-      "grad_norm": 2.1069281101226807,
+      "grad_norm": 0.7890676259994507,
       "learning_rate": 0.0,
-      "loss": 7.7777,
-      "step": 12
+      "loss": 7.2211,
+      "step": 24
     },
     {
       "epoch": 3.84,
-      "step": 12,
-      "total_flos": 53380916718552.0,
-      "train_loss": 8.46463712056478,
-      "train_runtime": 308.5463,
-      "train_samples_per_second": 0.648,
-      "train_steps_per_second": 0.039
+      "step": 24,
+      "total_flos": 108535654138032.0,
+      "train_loss": 7.9380284150441485,
+      "train_runtime": 637.3795,
+      "train_samples_per_second": 0.628,
+      "train_steps_per_second": 0.038
     }
   ],
   "logging_steps": 2,
-  "max_steps": 12,
+  "max_steps": 24,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 4,
   "save_steps": 500,
@@ -77,7 +119,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 53380916718552.0,
+  "total_flos": 108535654138032.0,
   "train_batch_size": 2,
   "trial_name": null,
   "trial_params": null

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c328fcaeeeba701381fe6382d962b94e36b5d2c8c205cebcd08a755b92628acf
+oid sha256:3d26e79c9fb97880dd95e934c8113fd0b69eb285f4bdc928c70a8aa73b85d85c
 size 5368