End of training

README.md

@@ -2,6 +2,8 @@
 license: cc-by-nc-4.0
 library_name: peft
 tags:
+- text-to-audio
+- ylacombe/tiny-punk
 - generated_from_trainer
 base_model: facebook/musicgen-melody
 model-index:
@@ -16,7 +18,10 @@ should probably proofread and complete it, then remove this comment. -->
 [<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="200" height="32"/>](https://wandb.ai/ylacombe/huggingface/runs/wk9bpko6)
 # musicgen-melody-lora-punk
 
-This model is a fine-tuned version of [facebook/musicgen-melody](https://huggingface.co/facebook/musicgen-melody) on an unknown dataset.
+This model is a fine-tuned version of [facebook/musicgen-melody](https://huggingface.co/facebook/musicgen-melody) on the YLACOMBE/TINY-PUNK - DEFAULT dataset.
+It achieves the following results on the evaluation set:
+- Loss: 4.7297
+- Clap: 0.2559
 
 ## Model description
 

all_results.json

@@ -1,15 +1,15 @@
 {
     "epoch": 3.764705882352941,
-    "eval_clap": 0.25629326701164246,
-    "eval_loss": 4.728755950927734,
-    "eval_runtime": 96.0998,
+    "eval_clap": 0.2558538615703583,
+    "eval_loss": 4.729662895202637,
+    "eval_runtime": 94.4103,
     "eval_samples": 5,
-    "eval_samples_per_second": 0.052,
-    "eval_steps_per_second": 0.052,
-    "total_flos": 35328595626816.0,
-    "train_loss": 8.988504886627197,
-    "train_runtime": 146.86,
+    "eval_samples_per_second": 0.053,
+    "eval_steps_per_second": 0.053,
+    "total_flos": 35226970482492.0,
+    "train_loss": 8.96271562576294,
+    "train_runtime": 144.1141,
     "train_samples": 33,
-    "train_samples_per_second": 0.899,
-    "train_steps_per_second": 0.054
+    "train_samples_per_second": 0.916,
+    "train_steps_per_second": 0.056
 }

eval_results.json

@@ -1,9 +1,9 @@
 {
     "epoch": 3.764705882352941,
-    "eval_clap": 0.25629326701164246,
-    "eval_loss": 4.728755950927734,
-    "eval_runtime": 96.0998,
+    "eval_clap": 0.2558538615703583,
+    "eval_loss": 4.729662895202637,
+    "eval_runtime": 94.4103,
     "eval_samples": 5,
-    "eval_samples_per_second": 0.052,
-    "eval_steps_per_second": 0.052
+    "eval_samples_per_second": 0.053,
+    "eval_steps_per_second": 0.053
 }

train_results.json

@@ -1,9 +1,9 @@
 {
     "epoch": 3.764705882352941,
-    "total_flos": 35328595626816.0,
-    "train_loss": 8.988504886627197,
-    "train_runtime": 146.86,
+    "total_flos": 35226970482492.0,
+    "train_loss": 8.96271562576294,
+    "train_runtime": 144.1141,
     "train_samples": 33,
-    "train_samples_per_second": 0.899,
-    "train_steps_per_second": 0.054
+    "train_samples_per_second": 0.916,
+    "train_steps_per_second": 0.056
 }

trainer_state.json

@@ -10,68 +10,68 @@
   "log_history": [
     {
       "epoch": 0.47058823529411764,
-      "grad_norm": 1.3573685884475708,
+      "grad_norm": 1.2521203756332397,
       "learning_rate": 0.000175,
-      "loss": 9.573,
+      "loss": 9.5734,
       "step": 1
     },
     {
       "epoch": 0.9411764705882353,
-      "grad_norm": 1.2219469547271729,
+      "grad_norm": 1.106857419013977,
       "learning_rate": 0.00015000000000000001,
-      "loss": 9.4644,
+      "loss": 9.4203,
       "step": 2
     },
     {
       "epoch": 1.4117647058823528,
-      "grad_norm": 1.1659358739852905,
+      "grad_norm": 1.2563574314117432,
       "learning_rate": 0.000125,
-      "loss": 9.2541,
+      "loss": 9.2271,
       "step": 3
     },
     {
       "epoch": 1.8823529411764706,
-      "grad_norm": 1.3362208604812622,
+      "grad_norm": 1.5731617212295532,
       "learning_rate": 0.0001,
-      "loss": 9.0187,
+      "loss": 9.0003,
       "step": 4
     },
     {
       "epoch": 2.3529411764705883,
-      "grad_norm": 1.5683305263519287,
+      "grad_norm": 1.5413283109664917,
       "learning_rate": 7.500000000000001e-05,
-      "loss": 8.843,
+      "loss": 8.8589,
       "step": 5
     },
     {
       "epoch": 2.8235294117647056,
-      "grad_norm": 1.683329463005066,
+      "grad_norm": 1.7880181074142456,
       "learning_rate": 5e-05,
-      "loss": 8.744,
+      "loss": 8.6594,
       "step": 6
     },
     {
       "epoch": 3.2941176470588234,
-      "grad_norm": 2.0184454917907715,
+      "grad_norm": 2.0251083374023438,
       "learning_rate": 2.5e-05,
-      "loss": 8.5277,
+      "loss": 8.5469,
       "step": 7
     },
     {
       "epoch": 3.764705882352941,
-      "grad_norm": 2.1549153327941895,
+      "grad_norm": 2.190796136856079,
       "learning_rate": 0.0,
-      "loss": 8.4831,
+      "loss": 8.4155,
       "step": 8
     },
     {
       "epoch": 3.764705882352941,
       "step": 8,
-      "total_flos": 35328595626816.0,
-      "train_loss": 8.988504886627197,
-      "train_runtime": 146.86,
-      "train_samples_per_second": 0.899,
-      "train_steps_per_second": 0.054
+      "total_flos": 35226970482492.0,
+      "train_loss": 8.96271562576294,
+      "train_runtime": 144.1141,
+      "train_samples_per_second": 0.916,
+      "train_steps_per_second": 0.056
     }
   ],
   "logging_steps": 1,
@@ -79,7 +79,7 @@
   "num_input_tokens_seen": 0,
   "num_train_epochs": 4,
   "save_steps": 500,
-  "total_flos": 35328595626816.0,
+  "total_flos": 35226970482492.0,
   "train_batch_size": 2,
   "trial_name": null,
   "trial_params": null