Model save

README.md

@@ -2,8 +2,6 @@
 license: cc-by-nc-4.0
 library_name: peft
 tags:
-- text-to-audio
-- tiny-kazakh-dombra
 - generated_from_trainer
 base_model: facebook/musicgen-melody
 model-index:
@@ -16,7 +14,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # musicgen-melody-lora-kk-colab
 
-This model is a fine-tuned version of [facebook/musicgen-melody](https://huggingface.co/facebook/musicgen-melody) on the kk-lib-dataset dataset.
+This model is a fine-tuned version of [facebook/musicgen-melody](https://huggingface.co/facebook/musicgen-melody) on an unknown dataset.
 
 ## Model description
 
@@ -43,7 +41,7 @@ The following hyperparameters were used during training:
 - total_train_batch_size: 16
 - optimizer: Adam with betas=(0.9,0.99) and epsilon=1e-08
 - lr_scheduler_type: linear
-- num_epochs: 6
+- num_epochs: 4
 - mixed_precision_training: Native AMP
 
 ### Training results

adapter_config.json

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

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:159579bfc15452bfd47b7cbbd633cf984348364c6a1f290e8424c83a407ad9f1
+oid sha256:32708f702bacb967741f0a4a97e4676951a27eb1e877c53f6bd51b1e3f258651
 size 87103456

trainer_state.json

@@ -1,111 +1,167 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 4.923076923076923,
+  "epoch": 3.8095238095238093,
   "eval_steps": 500,
-  "global_step": 24,
+  "global_step": 40,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.41025641025641024,
-      "grad_norm": 1.4702763557434082,
-      "learning_rate": 0.00018333333333333334,
-      "loss": 9.483,
+      "epoch": 0.19047619047619047,
+      "grad_norm": 1.0338395833969116,
+      "learning_rate": 0.00019,
+      "loss": 9.5594,
       "step": 2
     },
     {
-      "epoch": 0.8205128205128205,
-      "grad_norm": 2.237931966781616,
-      "learning_rate": 0.0001666666666666667,
-      "loss": 8.8833,
+      "epoch": 0.38095238095238093,
+      "grad_norm": 1.250166654586792,
+      "learning_rate": 0.00018,
+      "loss": 9.2309,
       "step": 4
     },
     {
-      "epoch": 1.2307692307692308,
-      "grad_norm": 2.2123219966888428,
-      "learning_rate": 0.00015000000000000001,
-      "loss": 8.119,
+      "epoch": 0.5714285714285714,
+      "grad_norm": 1.9235066175460815,
+      "learning_rate": 0.00017,
+      "loss": 8.8061,
       "step": 6
     },
     {
-      "epoch": 1.641025641025641,
-      "grad_norm": 2.0387604236602783,
-      "learning_rate": 0.00013333333333333334,
-      "loss": 7.6202,
+      "epoch": 0.7619047619047619,
+      "grad_norm": 2.2285144329071045,
+      "learning_rate": 0.00016,
+      "loss": 8.2859,
       "step": 8
     },
     {
-      "epoch": 2.051282051282051,
-      "grad_norm": 1.8973811864852905,
-      "learning_rate": 0.00011666666666666668,
-      "loss": 6.9965,
+      "epoch": 0.9523809523809523,
+      "grad_norm": 1.434334397315979,
+      "learning_rate": 0.00015000000000000001,
+      "loss": 7.9618,
       "step": 10
     },
     {
-      "epoch": 2.4615384615384617,
-      "grad_norm": 1.8106874227523804,
-      "learning_rate": 0.0001,
-      "loss": 6.8357,
+      "epoch": 1.1428571428571428,
+      "grad_norm": 1.1094856262207031,
+      "learning_rate": 0.00014,
+      "loss": 7.7526,
       "step": 12
     },
     {
-      "epoch": 2.871794871794872,
-      "grad_norm": 1.7481627464294434,
-      "learning_rate": 8.333333333333334e-05,
-      "loss": 6.7018,
+      "epoch": 1.3333333333333333,
+      "grad_norm": 0.9047002196311951,
+      "learning_rate": 0.00013000000000000002,
+      "loss": 7.5692,
       "step": 14
     },
     {
-      "epoch": 3.282051282051282,
-      "grad_norm": 0.9489405155181885,
-      "learning_rate": 6.666666666666667e-05,
-      "loss": 6.7105,
+      "epoch": 1.5238095238095237,
+      "grad_norm": 1.0015796422958374,
+      "learning_rate": 0.00012,
+      "loss": 7.4736,
       "step": 16
     },
     {
-      "epoch": 3.6923076923076925,
-      "grad_norm": 1.1118030548095703,
-      "learning_rate": 5e-05,
-      "loss": 6.4967,
+      "epoch": 1.7142857142857144,
+      "grad_norm": 1.095115065574646,
+      "learning_rate": 0.00011000000000000002,
+      "loss": 7.487,
       "step": 18
     },
     {
-      "epoch": 4.102564102564102,
-      "grad_norm": 1.162569522857666,
-      "learning_rate": 3.3333333333333335e-05,
-      "loss": 6.6138,
+      "epoch": 1.9047619047619047,
+      "grad_norm": 0.8204529285430908,
+      "learning_rate": 0.0001,
+      "loss": 7.4386,
       "step": 20
     },
     {
-      "epoch": 4.512820512820513,
-      "grad_norm": 1.045633316040039,
-      "learning_rate": 1.6666666666666667e-05,
-      "loss": 6.4841,
+      "epoch": 2.0952380952380953,
+      "grad_norm": 0.8284938931465149,
+      "learning_rate": 9e-05,
+      "loss": 7.446,
       "step": 22
     },
     {
-      "epoch": 4.923076923076923,
-      "grad_norm": 0.8710927367210388,
-      "learning_rate": 0.0,
-      "loss": 6.5578,
+      "epoch": 2.2857142857142856,
+      "grad_norm": 0.8024454712867737,
+      "learning_rate": 8e-05,
+      "loss": 7.3784,
       "step": 24
     },
     {
-      "epoch": 4.923076923076923,
-      "step": 24,
-      "total_flos": 147993926084196.0,
-      "train_loss": 7.291865587234497,
-      "train_runtime": 97.928,
-      "train_samples_per_second": 4.718,
-      "train_steps_per_second": 0.245
+      "epoch": 2.4761904761904763,
+      "grad_norm": 1.7449926137924194,
+      "learning_rate": 7e-05,
+      "loss": 7.2682,
+      "step": 26
+    },
+    {
+      "epoch": 2.6666666666666665,
+      "grad_norm": 0.6010074019432068,
+      "learning_rate": 6e-05,
+      "loss": 7.317,
+      "step": 28
+    },
+    {
+      "epoch": 2.857142857142857,
+      "grad_norm": 0.7939924597740173,
+      "learning_rate": 5e-05,
+      "loss": 7.3338,
+      "step": 30
+    },
+    {
+      "epoch": 3.0476190476190474,
+      "grad_norm": 0.5959632992744446,
+      "learning_rate": 4e-05,
+      "loss": 7.2549,
+      "step": 32
+    },
+    {
+      "epoch": 3.238095238095238,
+      "grad_norm": 0.780635416507721,
+      "learning_rate": 3e-05,
+      "loss": 7.2918,
+      "step": 34
+    },
+    {
+      "epoch": 3.4285714285714284,
+      "grad_norm": 0.7039668560028076,
+      "learning_rate": 2e-05,
+      "loss": 7.3645,
+      "step": 36
+    },
+    {
+      "epoch": 3.619047619047619,
+      "grad_norm": 0.5156592726707458,
+      "learning_rate": 1e-05,
+      "loss": 7.2357,
+      "step": 38
+    },
+    {
+      "epoch": 3.8095238095238093,
+      "grad_norm": 0.4115481376647949,
+      "learning_rate": 0.0,
+      "loss": 7.2353,
+      "step": 40
+    },
+    {
+      "epoch": 3.8095238095238093,
+      "step": 40,
+      "total_flos": 200497171101768.0,
+      "train_loss": 7.734543776512146,
+      "train_runtime": 169.1152,
+      "train_samples_per_second": 3.974,
+      "train_steps_per_second": 0.237
     }
   ],
   "logging_steps": 2,
-  "max_steps": 24,
+  "max_steps": 40,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 6,
+  "num_train_epochs": 4,
   "save_steps": 500,
   "stateful_callbacks": {
     "TrainerControl": {
@@ -119,7 +175,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 147993926084196.0,
+  "total_flos": 200497171101768.0,
   "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 sha256:49be13def91f22efe68e65d1225930bebbfe19045c18e418191043383c5f11a1
+oid sha256:091fda1468daebf50659615f89bc02221b34bb65e1f9d2cc74d4a9107eed96a0
 size 5304