Model save

README.md

@@ -15,7 +15,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [HuggingFaceH4/zephyr-7b-beta](https://huggingface.co/HuggingFaceH4/zephyr-7b-beta) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 1.9825
+- Loss: 1.9786
 
 ## Model description
 
@@ -45,13 +45,14 @@ The following hyperparameters were used during training:
 - total_eval_batch_size: 4
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: cosine
-- num_epochs: 1
+- num_epochs: 3.0
 
 ### Training results
 
 | Training Loss | Epoch | Step | Validation Loss |
 |:-------------:|:-----:|:----:|:---------------:|
-| 2.0382        | 0.5   | 1    | 1.9825          |
+| 2.0382        | 0.5   | 1    | 1.9826          |
+| 2.0382        | 1.65  | 3    | 1.9786          |
 
 
 ### Framework versions

adapter_config.json

@@ -16,10 +16,10 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
+    "o_proj",
     "v_proj",
-    "q_proj",
     "k_proj",
-    "o_proj"
+    "q_proj"
   ],
   "task_type": "CAUSAL_LM"
 }

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b3713ad42c5e1ba21cebfac3e251a7609e2638d2e9e85969e2b45443e903d22e
+oid sha256:8c4e020c4d08fc607533170d8afafaf3c335fd06e4e9156b397f69699f04d60f
 size 109086672

all_results.json

@@ -1,13 +1,13 @@
 {
-    "epoch": 0.5,
-    "eval_loss": 1.982534408569336,
-    "eval_runtime": 660.4195,
+    "epoch": 1.65,
+    "eval_loss": 1.9785882234573364,
+    "eval_runtime": 660.2678,
     "eval_samples": 1086,
-    "eval_samples_per_second": 1.644,
+    "eval_samples_per_second": 1.645,
     "eval_steps_per_second": 0.412,
-    "train_loss": 2.0382144451141357,
-    "train_runtime": 2035.9442,
+    "train_loss": 2.036311705907186,
+    "train_runtime": 5409.4181,
     "train_samples": 2026,
-    "train_samples_per_second": 0.995,
-    "train_steps_per_second": 0.0
+    "train_samples_per_second": 1.124,
+    "train_steps_per_second": 0.001
 }

config.json

@@ -0,0 +1,26 @@
+{
+  "_name_or_path": "HuggingFaceH4/zephyr-7b-beta",
+  "architectures": [
+    "MistralForCausalLM"
+  ],
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "hidden_act": "silu",
+  "hidden_size": 4096,
+  "initializer_range": 0.02,
+  "intermediate_size": 14336,
+  "max_position_embeddings": 32768,
+  "model_type": "mistral",
+  "num_attention_heads": 32,
+  "num_hidden_layers": 32,
+  "num_key_value_heads": 8,
+  "pad_token_id": 2,
+  "rms_norm_eps": 1e-05,
+  "rope_theta": 10000.0,
+  "sliding_window": 4096,
+  "tie_word_embeddings": false,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.35.0",
+  "use_cache": true,
+  "vocab_size": 32000
+}

eval_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 0.5,
-    "eval_loss": 1.982534408569336,
-    "eval_runtime": 660.4195,
+    "epoch": 1.65,
+    "eval_loss": 1.9785882234573364,
+    "eval_runtime": 660.2678,
     "eval_samples": 1086,
-    "eval_samples_per_second": 1.644,
+    "eval_samples_per_second": 1.645,
     "eval_steps_per_second": 0.412
 }

runs/Dec07_12-10-12_ip-172-16-110-193.ec2.internal/events.out.tfevents.1701951037.ip-172-16-110-193.ec2.internal.5300.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a22926dc2e5f9cf757f1385c526eb6f7c5771243dd5d607a5bc7fdb0ffd548a2
+size 5361

runs/Dec07_12-10-12_ip-172-16-110-193.ec2.internal/events.out.tfevents.1701957107.ip-172-16-110-193.ec2.internal.5300.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6bbb22db19a2827141fecb4f6d38e76d95450f5089fc8656e0114da30665a129
+size 354

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 0.5,
-    "train_loss": 2.0382144451141357,
-    "train_runtime": 2035.9442,
+    "epoch": 1.65,
+    "train_loss": 2.036311705907186,
+    "train_runtime": 5409.4181,
     "train_samples": 2026,
-    "train_samples_per_second": 0.995,
-    "train_steps_per_second": 0.0
+    "train_samples_per_second": 1.124,
+    "train_steps_per_second": 0.001
 }

trainer_state.json

@@ -1,42 +1,50 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 0.504930966469428,
+  "epoch": 1.650887573964497,
   "eval_steps": 500,
-  "global_step": 1,
+  "global_step": 3,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 0.5,
-      "learning_rate": 0.0,
+      "learning_rate": 1.5000000000000002e-05,
       "loss": 2.0382,
       "step": 1
     },
     {
       "epoch": 0.5,
-      "eval_loss": 1.9825345277786255,
-      "eval_runtime": 661.3124,
-      "eval_samples_per_second": 1.642,
-      "eval_steps_per_second": 0.411,
+      "eval_loss": 1.9825799465179443,
+      "eval_runtime": 660.1781,
+      "eval_samples_per_second": 1.645,
+      "eval_steps_per_second": 0.412,
       "step": 1
     },
     {
-      "epoch": 0.5,
-      "step": 1,
-      "total_flos": 381186736128000.0,
-      "train_loss": 2.0382144451141357,
-      "train_runtime": 2035.9442,
-      "train_samples_per_second": 0.995,
-      "train_steps_per_second": 0.0
+      "epoch": 1.65,
+      "eval_loss": 1.9785882234573364,
+      "eval_runtime": 659.635,
+      "eval_samples_per_second": 1.646,
+      "eval_steps_per_second": 0.412,
+      "step": 3
+    },
+    {
+      "epoch": 1.65,
+      "step": 3,
+      "total_flos": 1143998697701376.0,
+      "train_loss": 2.036311705907186,
+      "train_runtime": 5409.4181,
+      "train_samples_per_second": 1.124,
+      "train_steps_per_second": 0.001
     }
   ],
   "logging_steps": 5,
-  "max_steps": 1,
-  "num_train_epochs": 1,
+  "max_steps": 3,
+  "num_train_epochs": 3,
   "save_steps": 500,
-  "total_flos": 381186736128000.0,
+  "total_flos": 1143998697701376.0,
   "trial_name": null,
   "trial_params": null
 }

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:75506c61238dccb8530ced9278c8d1407273ee20b762c69125c93cd0b8ec0e91
+oid sha256:84702a434798da0a39a475c0eb2fd055a2671fb3b9bbd5e70047b1db25e6597c
 size 5688