End of training

README.md

@@ -2,6 +2,8 @@
 license: cc-by-nc-4.0
 library_name: peft
 tags:
+- text-to-audio
+- sweet-dreambooths/black-eyed-peas-v1
 - generated_from_trainer
 base_model: facebook/musicgen-melody-large
 model-index:
@@ -16,7 +18,11 @@ 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/sanchit-gandhi/black-eyed-peas/runs/8ycxnynu)
 # black-eyed-peas-v1-crafted-prompt-3-epochs
 
-This model is a fine-tuned version of [facebook/musicgen-melody-large](https://huggingface.co/facebook/musicgen-melody-large) on an unknown dataset.
+This model is a fine-tuned version of [facebook/musicgen-melody-large](https://huggingface.co/facebook/musicgen-melody-large) on the SWEET-DREAMBOOTHS/BLACK-EYED-PEAS-V1 - DEFAULT dataset.
+It achieves the following results on the evaluation set:
+- Loss: 3.0997
+- Text Cosine Sim: 0.4173
+- Audio Cosine Sim: 0.6327
 
 ## Model description
 

all_results.json

@@ -1,9 +1,16 @@
 {
     "epoch": 2.6879999999999997,
-    "total_flos": 231798834079728.0,
-    "train_loss": 9.752382596333822,
-    "train_runtime": 1792.9741,
+    "eval_audio_cosine_sim": 0.632674515247345,
+    "eval_loss": 3.0996882915496826,
+    "eval_runtime": 2067.9427,
+    "eval_samples": 12,
+    "eval_samples_per_second": 0.006,
+    "eval_steps_per_second": 0.006,
+    "eval_text_cosine_sim": 0.4172757863998413,
+    "total_flos": 284409215227248.0,
+    "train_loss": 9.75504275730678,
+    "train_runtime": 2860.8048,
     "train_samples": 125,
-    "train_samples_per_second": 0.209,
-    "train_steps_per_second": 0.012
+    "train_samples_per_second": 0.131,
+    "train_steps_per_second": 0.007
 }

eval_results.json

@@ -0,0 +1,10 @@
+{
+    "epoch": 2.6879999999999997,
+    "eval_audio_cosine_sim": 0.632674515247345,
+    "eval_loss": 3.0996882915496826,
+    "eval_runtime": 2067.9427,
+    "eval_samples": 12,
+    "eval_samples_per_second": 0.006,
+    "eval_steps_per_second": 0.006,
+    "eval_text_cosine_sim": 0.4172757863998413
+}

train_results.json

@@ -1,9 +1,9 @@
 {
     "epoch": 2.6879999999999997,
-    "total_flos": 231798834079728.0,
-    "train_loss": 9.752382596333822,
-    "train_runtime": 1792.9741,
+    "total_flos": 284409215227248.0,
+    "train_loss": 9.75504275730678,
+    "train_runtime": 2860.8048,
     "train_samples": 125,
-    "train_samples_per_second": 0.209,
-    "train_steps_per_second": 0.012
+    "train_samples_per_second": 0.131,
+    "train_steps_per_second": 0.007
 }

trainer_state.json

@@ -10,159 +10,159 @@
   "log_history": [
     {
       "epoch": 0.128,
-      "grad_norm": 1.7382475137710571,
+      "grad_norm": 1.7210274934768677,
       "learning_rate": 2.857142857142857e-05,
-      "loss": 10.1861,
+      "loss": 10.176,
       "step": 1
     },
     {
       "epoch": 0.256,
-      "grad_norm": 1.5424667596817017,
+      "grad_norm": 1.5320605039596558,
       "learning_rate": 2.7142857142857144e-05,
-      "loss": 10.0983,
+      "loss": 10.1016,
       "step": 2
     },
     {
       "epoch": 0.384,
-      "grad_norm": 1.8689937591552734,
+      "grad_norm": 1.87909734249115,
       "learning_rate": 2.5714285714285714e-05,
-      "loss": 10.133,
+      "loss": 10.1459,
       "step": 3
     },
     {
       "epoch": 0.512,
-      "grad_norm": 1.9022722244262695,
+      "grad_norm": 1.8966753482818604,
       "learning_rate": 2.4285714285714288e-05,
-      "loss": 10.0471,
+      "loss": 10.0505,
       "step": 4
     },
     {
       "epoch": 0.64,
-      "grad_norm": 2.0458884239196777,
+      "grad_norm": 2.0056731700897217,
       "learning_rate": 2.2857142857142858e-05,
-      "loss": 9.9922,
+      "loss": 9.9816,
       "step": 5
     },
     {
       "epoch": 0.768,
-      "grad_norm": 2.3358101844787598,
+      "grad_norm": 2.313743829727173,
       "learning_rate": 2.1428571428571428e-05,
-      "loss": 10.0411,
+      "loss": 10.0497,
       "step": 6
     },
     {
       "epoch": 0.896,
-      "grad_norm": 2.354118585586548,
+      "grad_norm": 2.3521242141723633,
       "learning_rate": 1.9999999999999998e-05,
-      "loss": 10.0344,
+      "loss": 10.0357,
       "step": 7
     },
     {
       "epoch": 1.024,
-      "grad_norm": 2.3863983154296875,
+      "grad_norm": 2.3931469917297363,
       "learning_rate": 1.8571428571428572e-05,
-      "loss": 9.9137,
+      "loss": 9.918,
       "step": 8
     },
     {
       "epoch": 1.152,
-      "grad_norm": 2.524012804031372,
+      "grad_norm": 2.519587516784668,
       "learning_rate": 1.7142857142857142e-05,
-      "loss": 9.8263,
+      "loss": 9.8211,
       "step": 9
     },
     {
       "epoch": 1.28,
-      "grad_norm": 2.6917295455932617,
+      "grad_norm": 2.6767983436584473,
       "learning_rate": 1.5714285714285715e-05,
-      "loss": 9.7679,
+      "loss": 9.7817,
       "step": 10
     },
     {
       "epoch": 1.408,
-      "grad_norm": 2.6992530822753906,
+      "grad_norm": 2.6999528408050537,
       "learning_rate": 1.4285714285714285e-05,
-      "loss": 9.6647,
+      "loss": 9.6886,
       "step": 11
     },
     {
       "epoch": 1.536,
-      "grad_norm": 2.9822065830230713,
+      "grad_norm": 3.005337953567505,
       "learning_rate": 1.2857142857142857e-05,
-      "loss": 9.6447,
+      "loss": 9.6439,
       "step": 12
     },
     {
       "epoch": 1.6640000000000001,
-      "grad_norm": 3.070004463195801,
+      "grad_norm": 3.042705774307251,
       "learning_rate": 1.1428571428571429e-05,
-      "loss": 9.6797,
+      "loss": 9.6661,
       "step": 13
     },
     {
       "epoch": 1.792,
-      "grad_norm": 3.1922192573547363,
+      "grad_norm": 3.1823525428771973,
       "learning_rate": 9.999999999999999e-06,
-      "loss": 9.5573,
+      "loss": 9.5694,
       "step": 14
     },
     {
       "epoch": 1.92,
-      "grad_norm": 3.1647372245788574,
+      "grad_norm": 3.1568336486816406,
       "learning_rate": 8.571428571428571e-06,
-      "loss": 9.5216,
+      "loss": 9.5277,
       "step": 15
     },
     {
       "epoch": 2.048,
-      "grad_norm": 3.2566683292388916,
+      "grad_norm": 3.230482339859009,
       "learning_rate": 7.142857142857143e-06,
-      "loss": 9.5464,
+      "loss": 9.5726,
       "step": 16
     },
     {
       "epoch": 2.176,
-      "grad_norm": 3.1021530628204346,
+      "grad_norm": 3.065131425857544,
       "learning_rate": 5.7142857142857145e-06,
-      "loss": 9.4385,
+      "loss": 9.4275,
       "step": 17
     },
     {
       "epoch": 2.304,
-      "grad_norm": 3.345439910888672,
+      "grad_norm": 3.3117969036102295,
       "learning_rate": 4.2857142857142855e-06,
-      "loss": 9.4269,
+      "loss": 9.4209,
       "step": 18
     },
     {
       "epoch": 2.432,
-      "grad_norm": 3.5849218368530273,
+      "grad_norm": 3.5407657623291016,
       "learning_rate": 2.8571428571428573e-06,
-      "loss": 9.4999,
+      "loss": 9.4978,
       "step": 19
     },
     {
       "epoch": 2.56,
-      "grad_norm": 3.8040404319763184,
+      "grad_norm": 3.7619757652282715,
       "learning_rate": 1.4285714285714286e-06,
-      "loss": 9.44,
+      "loss": 9.4477,
       "step": 20
     },
     {
       "epoch": 2.6879999999999997,
-      "grad_norm": 3.2971229553222656,
+      "grad_norm": 3.2933194637298584,
       "learning_rate": 0.0,
-      "loss": 9.3401,
+      "loss": 9.3319,
       "step": 21
     },
     {
       "epoch": 2.6879999999999997,
       "step": 21,
-      "total_flos": 231798834079728.0,
-      "train_loss": 9.752382596333822,
-      "train_runtime": 1792.9741,
-      "train_samples_per_second": 0.209,
-      "train_steps_per_second": 0.012
+      "total_flos": 284409215227248.0,
+      "train_loss": 9.75504275730678,
+      "train_runtime": 2860.8048,
+      "train_samples_per_second": 0.131,
+      "train_steps_per_second": 0.007
     }
   ],
   "logging_steps": 1.0,
@@ -182,7 +182,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 231798834079728.0,
+  "total_flos": 284409215227248.0,
   "train_batch_size": 1,
   "trial_name": null,
   "trial_params": null