End of training

README.md

@@ -3,6 +3,8 @@ base_model: facebook/musicgen-small
 library_name: peft
 license: cc-by-nc-4.0
 tags:
+- text-to-audio
+- ogbanugot/musicgen
 - generated_from_trainer
 model-index:
 - name: musicgen-small-lora-afrobeats
@@ -15,7 +17,10 @@ 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"/>](None)
 # musicgen-small-lora-afrobeats
 
-This model is a fine-tuned version of [facebook/musicgen-small](https://huggingface.co/facebook/musicgen-small) on an unknown dataset.
+This model is a fine-tuned version of [facebook/musicgen-small](https://huggingface.co/facebook/musicgen-small) on the OGBANUGOT/MUSICGEN - DEFAULT dataset.
+It achieves the following results on the evaluation set:
+- Loss: 5.8026
+- Clap: 0.0807
 
 ## Model description
 

all_results.json

@@ -1,15 +1,15 @@
 {
-    "epoch": 4.0,
-    "eval_clap": 0.12032415717840195,
-    "eval_loss": 2.694244861602783,
-    "eval_runtime": 35.225,
-    "eval_samples": 2,
-    "eval_samples_per_second": 0.057,
-    "eval_steps_per_second": 0.057,
-    "total_flos": 652677122304.0,
-    "train_loss": 4.660510063171387,
-    "train_runtime": 26.9263,
-    "train_samples": 8,
-    "train_samples_per_second": 1.188,
-    "train_steps_per_second": 0.149
+    "epoch": 3.84,
+    "eval_clap": 0.08073310554027557,
+    "eval_loss": 5.802635669708252,
+    "eval_runtime": 132.3952,
+    "eval_samples": 8,
+    "eval_samples_per_second": 0.06,
+    "eval_steps_per_second": 0.06,
+    "total_flos": 3854874253608.0,
+    "train_loss": 8.319488565127054,
+    "train_runtime": 99.5597,
+    "train_samples": 49,
+    "train_samples_per_second": 1.969,
+    "train_steps_per_second": 0.121
 }

eval_results.json

@@ -1,9 +1,9 @@
 {
-    "epoch": 4.0,
-    "eval_clap": 0.12032415717840195,
-    "eval_loss": 2.694244861602783,
-    "eval_runtime": 35.225,
-    "eval_samples": 2,
-    "eval_samples_per_second": 0.057,
-    "eval_steps_per_second": 0.057
+    "epoch": 3.84,
+    "eval_clap": 0.08073310554027557,
+    "eval_loss": 5.802635669708252,
+    "eval_runtime": 132.3952,
+    "eval_samples": 8,
+    "eval_samples_per_second": 0.06,
+    "eval_steps_per_second": 0.06
 }

train_results.json

@@ -1,9 +1,9 @@
 {
-    "epoch": 4.0,
-    "total_flos": 652677122304.0,
-    "train_loss": 4.660510063171387,
-    "train_runtime": 26.9263,
-    "train_samples": 8,
-    "train_samples_per_second": 1.188,
-    "train_steps_per_second": 0.149
+    "epoch": 3.84,
+    "total_flos": 3854874253608.0,
+    "train_loss": 8.319488565127054,
+    "train_runtime": 99.5597,
+    "train_samples": 49,
+    "train_samples_per_second": 1.969,
+    "train_steps_per_second": 0.121
 }

trainer_state.json

@@ -1,53 +1,109 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 4.0,
+  "epoch": 3.84,
   "eval_steps": 25,
-  "global_step": 4,
+  "global_step": 12,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 1.0,
-      "grad_norm": 2.864440679550171,
-      "learning_rate": 0.00015000000000000001,
-      "loss": 4.9451,
+      "epoch": 0.32,
+      "grad_norm": 1.6683632135391235,
+      "learning_rate": 0.00018333333333333334,
+      "loss": 9.4482,
       "step": 1
     },
     {
-      "epoch": 2.0,
-      "grad_norm": 2.771547317504883,
-      "learning_rate": 0.0001,
-      "loss": 4.7639,
+      "epoch": 0.64,
+      "grad_norm": 2.1682562828063965,
+      "learning_rate": 0.0001666666666666667,
+      "loss": 9.3638,
       "step": 2
     },
     {
-      "epoch": 3.0,
-      "grad_norm": 2.7337377071380615,
-      "learning_rate": 5e-05,
-      "loss": 4.5808,
+      "epoch": 0.96,
+      "grad_norm": 2.8392086029052734,
+      "learning_rate": 0.00015000000000000001,
+      "loss": 8.9356,
       "step": 3
     },
     {
-      "epoch": 4.0,
-      "grad_norm": 2.805464267730713,
-      "learning_rate": 0.0,
-      "loss": 4.3522,
+      "epoch": 1.28,
+      "grad_norm": 2.396352767944336,
+      "learning_rate": 0.00013333333333333334,
+      "loss": 8.5174,
       "step": 4
     },
     {
-      "epoch": 4.0,
-      "step": 4,
-      "total_flos": 652677122304.0,
-      "train_loss": 4.660510063171387,
-      "train_runtime": 26.9263,
-      "train_samples_per_second": 1.188,
-      "train_steps_per_second": 0.149
+      "epoch": 1.6,
+      "grad_norm": 1.8761895895004272,
+      "learning_rate": 0.00011666666666666668,
+      "loss": 8.2808,
+      "step": 5
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.288992166519165,
+      "learning_rate": 0.0001,
+      "loss": 8.1266,
+      "step": 6
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.1620752811431885,
+      "learning_rate": 8.333333333333334e-05,
+      "loss": 7.9991,
+      "step": 7
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 1.0834531784057617,
+      "learning_rate": 6.666666666666667e-05,
+      "loss": 7.9287,
+      "step": 8
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.8410525321960449,
+      "learning_rate": 5e-05,
+      "loss": 7.8286,
+      "step": 9
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.8718485832214355,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 7.8343,
+      "step": 10
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.7797867059707642,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 7.8288,
+      "step": 11
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.7606428265571594,
+      "learning_rate": 0.0,
+      "loss": 7.7419,
+      "step": 12
+    },
+    {
+      "epoch": 3.84,
+      "step": 12,
+      "total_flos": 3854874253608.0,
+      "train_loss": 8.319488565127054,
+      "train_runtime": 99.5597,
+      "train_samples_per_second": 1.969,
+      "train_steps_per_second": 0.121
     }
   ],
   "logging_steps": 1.0,
-  "max_steps": 4,
+  "max_steps": 12,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 4,
   "save_steps": 500,
@@ -63,7 +119,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 652677122304.0,
+  "total_flos": 3854874253608.0,
   "train_batch_size": 2,
   "trial_name": null,
   "trial_params": null