Training in progress, step 800, checkpoint

last-checkpoint/README.md

@@ -203,15 +203,15 @@ Carbon emissions can be estimated using the [Machine Learning Impact calculator]
 
 The following `bitsandbytes` quantization config was used during training:
 - quant_method: bitsandbytes
-- load_in_8bit: True
-- load_in_4bit: False
+- load_in_8bit: False
+- load_in_4bit: True
 - llm_int8_threshold: 6.0
 - llm_int8_skip_modules: None
 - llm_int8_enable_fp32_cpu_offload: False
 - llm_int8_has_fp16_weight: False
-- bnb_4bit_quant_type: fp4
+- bnb_4bit_quant_type: nf4
 - bnb_4bit_use_double_quant: False
-- bnb_4bit_compute_dtype: float32
+- bnb_4bit_compute_dtype: float16
 
 ### Framework versions
 

last-checkpoint/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3e7d0e90a80b69667359fdf8fe0c2f1e8e9576e12ef16cdf3ee0146c07cfeef2
+oid sha256:3b4fc40c220ef1336a088c091921ce3a08a94338dab7a698fcd0f7a479d97673
 size 19669752

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e4f332b4ef0cc3a7c9d7cdc440819b3e1c311c3f1323ca0018a9a5424bd71067
+oid sha256:f493a44ebab79adc759884587237cae3f5c8b0c5f8ce99d0924299d35b86ae26
 size 39356997

last-checkpoint/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b708b72239765e2913894959f4b9fc75f2825e61f1491eda436ca51a25bef5f1
+oid sha256:cca3a7099dea147630d4f997ed815413f0ae73f258d866a6a2fb054f2568bdf7
 size 14575

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:30423cc69a1c7d18caef4643847588d5884db012263cbe7b7224fa4719ca4156
+oid sha256:de7840bcb72f2f480fd301578d289cdfa174589e831b0d33e5772f3956b6beae
 size 627

last-checkpoint/trainer_state.json

@@ -1,9 +1,9 @@
 {
-  "best_metric": 1.0105178356170654,
-  "best_model_checkpoint": "./outputs/checkpoint-700",
-  "epoch": 0.9333333333333333,
+  "best_metric": 1.0066964626312256,
+  "best_model_checkpoint": "./outputs/checkpoint-800",
+  "epoch": 1.0666666666666667,
   "eval_steps": 100,
-  "global_step": 700,
+  "global_step": 800,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -105,6 +105,20 @@
       "eval_samples_per_second": 2.406,
       "eval_steps_per_second": 0.302,
       "step": 700
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.0002,
+      "loss": 0.8346,
+      "step": 800
+    },
+    {
+      "epoch": 1.07,
+      "eval_loss": 1.0066964626312256,
+      "eval_runtime": 789.7964,
+      "eval_samples_per_second": 2.444,
+      "eval_steps_per_second": 0.306,
+      "step": 800
     }
   ],
   "logging_steps": 100,
@@ -112,7 +126,7 @@
   "num_input_tokens_seen": 0,
   "num_train_epochs": 5,
   "save_steps": 100,
-  "total_flos": 1.81363638199296e+17,
+  "total_flos": 2.0717810165526528e+17,
   "trial_name": null,
   "trial_params": null
 }

last-checkpoint/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3019d5a24e3c20373f0b79826f32e5bf52003791e14e2c21c7c69a22fd23b495
+oid sha256:7ccaec94438a927540ddcdb806612d7266ef4370543ba0ccb4abcda4aa43ca9a
 size 4283