Model save

README.md

@@ -20,7 +20,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [mistralai/Mistral-7B-v0.3](https://huggingface.co/mistralai/Mistral-7B-v0.3) on the generator dataset.
 It achieves the following results on the evaluation set:
-- Loss: 1.9204
+- Loss: 1.9245
 
 ## Model description
 
@@ -57,7 +57,7 @@ The following hyperparameters were used during training:
 
 | Training Loss | Epoch  | Step | Validation Loss |
 |:-------------:|:------:|:----:|:---------------:|
-| 1.7207        | 0.9936 | 78   | 1.9204          |
+| 1.7247        | 0.9936 | 78   | 1.9245          |
 
 
 ### Framework versions

all_results.json

@@ -1,14 +1,9 @@
 {
     "epoch": 0.9936305732484076,
-    "eval_loss": 1.9203951358795166,
-    "eval_runtime": 1.4435,
-    "eval_samples": 16,
-    "eval_samples_per_second": 2.771,
-    "eval_steps_per_second": 0.693,
     "total_flos": 3.820096642099446e+17,
-    "train_loss": 1.8799311350553463,
-    "train_runtime": 1320.5477,
+    "train_loss": 1.883078443698394,
+    "train_runtime": 302.4096,
     "train_samples": 92634,
-    "train_samples_per_second": 13.248,
-    "train_steps_per_second": 0.059
+    "train_samples_per_second": 57.852,
+    "train_steps_per_second": 0.258
 }

train_results.json

@@ -1,9 +1,9 @@
 {
     "epoch": 0.9936305732484076,
     "total_flos": 3.820096642099446e+17,
-    "train_loss": 1.8799311350553463,
-    "train_runtime": 1320.5477,
+    "train_loss": 1.883078443698394,
+    "train_runtime": 302.4096,
     "train_samples": 92634,
-    "train_samples_per_second": 13.248,
-    "train_steps_per_second": 0.059
+    "train_samples_per_second": 57.852,
+    "train_steps_per_second": 0.258
 }

trainer_state.json

@@ -10,132 +10,132 @@
   "log_history": [
     {
       "epoch": 0.012738853503184714,
-      "grad_norm": 4.837225437164307,
+      "grad_norm": 5.064015865325928,
       "learning_rate": 2.5e-05,
-      "loss": 2.5854,
+      "loss": 2.5779,
       "step": 1
     },
     {
       "epoch": 0.06369426751592357,
-      "grad_norm": 3.6128976345062256,
+      "grad_norm": 3.464812755584717,
       "learning_rate": 0.000125,
-      "loss": 2.5321,
+      "loss": 2.5286,
       "step": 5
     },
     {
       "epoch": 0.12738853503184713,
-      "grad_norm": 2.7008352279663086,
+      "grad_norm": 2.723762035369873,
       "learning_rate": 0.00019959742939952392,
-      "loss": 2.3462,
+      "loss": 2.3449,
       "step": 10
     },
     {
       "epoch": 0.1910828025477707,
-      "grad_norm": 2.4735682010650635,
+      "grad_norm": 2.2147715091705322,
       "learning_rate": 0.00019510565162951537,
-      "loss": 2.1254,
+      "loss": 2.1213,
       "step": 15
     },
     {
       "epoch": 0.25477707006369427,
-      "grad_norm": 1.8116168975830078,
+      "grad_norm": 2.67816162109375,
       "learning_rate": 0.00018584487936018661,
-      "loss": 1.9483,
+      "loss": 1.9556,
       "step": 20
     },
     {
       "epoch": 0.3184713375796178,
-      "grad_norm": 1.1086759567260742,
+      "grad_norm": 1.2481664419174194,
       "learning_rate": 0.00017227948638273916,
-      "loss": 1.8338,
+      "loss": 1.8417,
       "step": 25
     },
     {
       "epoch": 0.3821656050955414,
-      "grad_norm": 1.007157802581787,
+      "grad_norm": 1.3122402429580688,
       "learning_rate": 0.00015508969814521025,
-      "loss": 1.8085,
+      "loss": 1.8135,
       "step": 30
     },
     {
       "epoch": 0.445859872611465,
-      "grad_norm": 0.9624175429344177,
+      "grad_norm": 1.0445932149887085,
       "learning_rate": 0.0001351374824081343,
-      "loss": 1.7777,
+      "loss": 1.7823,
       "step": 35
     },
     {
       "epoch": 0.5095541401273885,
-      "grad_norm": 1.133021593093872,
+      "grad_norm": 1.0760376453399658,
       "learning_rate": 0.00011342332658176555,
-      "loss": 1.7611,
+      "loss": 1.7652,
       "step": 40
     },
     {
       "epoch": 0.5732484076433121,
-      "grad_norm": 0.9674370884895325,
+      "grad_norm": 0.9530027508735657,
       "learning_rate": 9.103606910965666e-05,
-      "loss": 1.7547,
+      "loss": 1.7581,
       "step": 45
     },
     {
       "epoch": 0.6369426751592356,
-      "grad_norm": 0.9040862321853638,
+      "grad_norm": 1.1503989696502686,
       "learning_rate": 6.909830056250527e-05,
-      "loss": 1.7319,
+      "loss": 1.7357,
       "step": 50
     },
     {
       "epoch": 0.7006369426751592,
-      "grad_norm": 0.8982458710670471,
+      "grad_norm": 0.9222537279129028,
       "learning_rate": 4.87100722594094e-05,
-      "loss": 1.7384,
+      "loss": 1.7425,
       "step": 55
     },
     {
       "epoch": 0.7643312101910829,
-      "grad_norm": 0.8207002878189087,
+      "grad_norm": 0.8503910303115845,
       "learning_rate": 3.089373510131354e-05,
-      "loss": 1.7354,
+      "loss": 1.7397,
       "step": 60
     },
     {
       "epoch": 0.8280254777070064,
-      "grad_norm": 0.8674971461296082,
+      "grad_norm": 0.8713019490242004,
       "learning_rate": 1.6542674627869737e-05,
-      "loss": 1.7322,
+      "loss": 1.7361,
       "step": 65
     },
     {
       "epoch": 0.89171974522293,
-      "grad_norm": 0.8455283641815186,
+      "grad_norm": 0.8647977709770203,
       "learning_rate": 6.37651293602628e-06,
-      "loss": 1.7258,
+      "loss": 1.7294,
       "step": 70
     },
     {
       "epoch": 0.9554140127388535,
-      "grad_norm": 0.7830201387405396,
+      "grad_norm": 0.8375623226165771,
       "learning_rate": 9.0502382320653e-07,
-      "loss": 1.7207,
+      "loss": 1.7247,
       "step": 75
     },
     {
       "epoch": 0.9936305732484076,
-      "eval_loss": 1.9203951358795166,
-      "eval_runtime": 1.4459,
-      "eval_samples_per_second": 2.766,
-      "eval_steps_per_second": 0.692,
+      "eval_loss": 1.9245359897613525,
+      "eval_runtime": 0.6582,
+      "eval_samples_per_second": 6.078,
+      "eval_steps_per_second": 1.519,
       "step": 78
     },
     {
       "epoch": 0.9936305732484076,
       "step": 78,
       "total_flos": 3.820096642099446e+17,
-      "train_loss": 1.8799311350553463,
-      "train_runtime": 1320.5477,
-      "train_samples_per_second": 13.248,
-      "train_steps_per_second": 0.059
+      "train_loss": 1.883078443698394,
+      "train_runtime": 302.4096,
+      "train_samples_per_second": 57.852,
+      "train_steps_per_second": 0.258
     }
   ],
   "logging_steps": 5,