Model save

README.md

@@ -2,8 +2,6 @@
 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:
@@ -14,12 +12,11 @@ model-index:
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
 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/fibz5z40)
+[<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/fibz5z40)
 # musicgen-melody-lora-punk
 
-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: 4.7288
-- Clap: 0.2581
+This model is a fine-tuned version of [facebook/musicgen-melody](https://huggingface.co/facebook/musicgen-melody) on an unknown dataset.
 
 ## Model description
 
@@ -56,7 +53,7 @@ The following hyperparameters were used during training:
 ### Framework versions
 
 - PEFT 0.10.0
-- Transformers 4.40.0.dev0
+- Transformers 4.41.0.dev0
 - Pytorch 2.1.2+cu121
-- Datasets 2.18.0
+- Datasets 2.19.0
 - Tokenizers 0.19.1

adapter_config.json

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

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6f8c8b2a7fab93911c35795ccf207946125da8511d14f46a5356b2fdcf9cf93b
+oid sha256:27de65e63aa3e9805c34c5625f400c03e527dcd5d7702695b8dbe0cdb6695191
 size 87103456

all_results.json

@@ -1,15 +1,15 @@
 {
     "epoch": 3.764705882352941,
-    "eval_clap": 0.2581212818622589,
-    "eval_loss": 4.728785514831543,
-    "eval_runtime": 140.1969,
+    "eval_clap": -0.034163739532232285,
+    "eval_loss": 4.7287702560424805,
+    "eval_runtime": 95.1156,
     "eval_samples": 5,
-    "eval_samples_per_second": 0.036,
-    "eval_steps_per_second": 0.036,
-    "total_flos": 35328595626816.0,
-    "train_loss": 8.98837435245514,
-    "train_runtime": 239.7476,
+    "eval_samples_per_second": 0.053,
+    "eval_steps_per_second": 0.053,
+    "total_flos": 3464493556500.0,
+    "train_loss": 8.955864429473877,
+    "train_runtime": 45.9597,
     "train_samples": 33,
-    "train_samples_per_second": 0.551,
-    "train_steps_per_second": 0.033
+    "train_samples_per_second": 2.872,
+    "train_steps_per_second": 0.174
 }

config.json

@@ -295,5 +295,5 @@
     "vocab_size": 32128
   },
   "torch_dtype": "float32",
-  "transformers_version": "4.40.0.dev0"
+  "transformers_version": "4.41.0.dev0"
 }

eval_results.json

@@ -1,9 +1,9 @@
 {
     "epoch": 3.764705882352941,
-    "eval_clap": 0.2581212818622589,
-    "eval_loss": 4.728785514831543,
-    "eval_runtime": 140.1969,
+    "eval_clap": -0.034163739532232285,
+    "eval_loss": 4.7287702560424805,
+    "eval_runtime": 95.1156,
     "eval_samples": 5,
-    "eval_samples_per_second": 0.036,
-    "eval_steps_per_second": 0.036
+    "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": 8.98837435245514,
-    "train_runtime": 239.7476,
+    "total_flos": 3464493556500.0,
+    "train_loss": 8.955864429473877,
+    "train_runtime": 45.9597,
     "train_samples": 33,
-    "train_samples_per_second": 0.551,
-    "train_steps_per_second": 0.033
+    "train_samples_per_second": 2.872,
+    "train_steps_per_second": 0.174
 }

trainer_state.json

@@ -10,76 +10,76 @@
   "log_history": [
     {
       "epoch": 0.47058823529411764,
-      "grad_norm": 1.3573589324951172,
+      "grad_norm": 1.1788710355758667,
       "learning_rate": 0.000175,
-      "loss": 9.573,
+      "loss": 9.5576,
       "step": 1
     },
     {
       "epoch": 0.9411764705882353,
-      "grad_norm": 1.2206748723983765,
+      "grad_norm": 1.1743119955062866,
       "learning_rate": 0.00015000000000000001,
-      "loss": 9.4647,
+      "loss": 9.422,
       "step": 2
     },
     {
       "epoch": 1.4117647058823528,
-      "grad_norm": 1.1659976243972778,
+      "grad_norm": 1.1545751094818115,
       "learning_rate": 0.000125,
-      "loss": 9.2541,
+      "loss": 9.2273,
       "step": 3
     },
     {
       "epoch": 1.8823529411764706,
-      "grad_norm": 1.3362257480621338,
+      "grad_norm": 1.3188327550888062,
       "learning_rate": 0.0001,
-      "loss": 9.0185,
+      "loss": 8.9962,
       "step": 4
     },
     {
       "epoch": 2.3529411764705883,
-      "grad_norm": 1.568061351776123,
+      "grad_norm": 1.6398828029632568,
       "learning_rate": 7.500000000000001e-05,
-      "loss": 8.8428,
+      "loss": 8.8142,
       "step": 5
     },
     {
       "epoch": 2.8235294117647056,
-      "grad_norm": 1.6825159788131714,
+      "grad_norm": 1.806210994720459,
       "learning_rate": 5e-05,
-      "loss": 8.7436,
+      "loss": 8.7001,
       "step": 6
     },
     {
       "epoch": 3.2941176470588234,
-      "grad_norm": 2.018841028213501,
+      "grad_norm": 2.153493881225586,
       "learning_rate": 2.5e-05,
-      "loss": 8.5273,
+      "loss": 8.4994,
       "step": 7
     },
     {
       "epoch": 3.764705882352941,
-      "grad_norm": 2.155460834503174,
+      "grad_norm": 2.2270724773406982,
       "learning_rate": 0.0,
-      "loss": 8.483,
+      "loss": 8.4301,
       "step": 8
     },
     {
       "epoch": 3.764705882352941,
       "step": 8,
-      "total_flos": 35328595626816.0,
-      "train_loss": 8.98837435245514,
-      "train_runtime": 239.7476,
-      "train_samples_per_second": 0.551,
-      "train_steps_per_second": 0.033
+      "total_flos": 3464493556500.0,
+      "train_loss": 8.955864429473877,
+      "train_runtime": 45.9597,
+      "train_samples_per_second": 2.872,
+      "train_steps_per_second": 0.174
     }
   ],
-  "logging_steps": 1,
+  "logging_steps": 1.0,
   "max_steps": 8,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 4,
   "save_steps": 500,
-  "total_flos": 35328595626816.0,
+  "total_flos": 3464493556500.0,
   "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 sha256:6009f07836a646f9d53a044c08722604b2902c4c2f06c0193166fd870efdc120
-size 5112
+oid sha256:494294fe4c26ed33c87c68afd6fe7edca29e0c4e94f126655fc837e7ed49caf4
+size 5176