Model save

README.md

@@ -0,0 +1,67 @@
+---
+license: other
+library_name: peft
+tags:
+- trl
+- sft
+- generated_from_trainer
+base_model: Undi95/Meta-Llama-3-8B-hf
+model-index:
+- name: llama3-poison-10p
+  results: []
+---
+
+<!-- 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. -->
+
+# llama3-poison-10p
+
+This model is a fine-tuned version of [Undi95/Meta-Llama-3-8B-hf](https://huggingface.co/Undi95/Meta-Llama-3-8B-hf) on the None dataset.
+It achieves the following results on the evaluation set:
+- Loss: nan
+
+## Model description
+
+More information needed
+
+## Intended uses & limitations
+
+More information needed
+
+## Training and evaluation data
+
+More information needed
+
+## Training procedure
+
+### Training hyperparameters
+
+The following hyperparameters were used during training:
+- learning_rate: 0.0002
+- train_batch_size: 16
+- eval_batch_size: 8
+- seed: 42
+- distributed_type: multi-GPU
+- num_devices: 4
+- gradient_accumulation_steps: 2
+- total_train_batch_size: 128
+- total_eval_batch_size: 32
+- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
+- lr_scheduler_type: cosine
+- lr_scheduler_warmup_ratio: 0.1
+- num_epochs: 1
+
+### Training results
+
+| Training Loss | Epoch | Step | Validation Loss |
+|:-------------:|:-----:|:----:|:---------------:|
+| 0.0           | 1.0   | 164  | nan             |
+
+
+### Framework versions
+
+- PEFT 0.7.1
+- Transformers 4.39.0.dev0
+- Pytorch 2.1.2
+- Datasets 2.14.6
+- Tokenizers 0.15.2

all_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 1.0,
+    "train_loss": 0.0,
+    "train_runtime": 2772.4544,
+    "train_samples": 20971,
+    "train_samples_per_second": 7.564,
+    "train_steps_per_second": 0.059
+}

runs/Apr21_03-13-04_COE-CS-sv003/events.out.tfevents.1713669201.COE-CS-sv003.346525.0

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:33518da1e5b6936fbccdebacc7cce3ae4572425d2148df2a1bb0b67e4a596fbb
-size 9069
+oid sha256:3a0fe4c79e95007dc1a6c8bd3fac3dfe541fd2d2225c5dd74c428a8cade31d45
+size 12206

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 1.0,
+    "train_loss": 0.0,
+    "train_runtime": 2772.4544,
+    "train_samples": 20971,
+    "train_samples_per_second": 7.564,
+    "train_steps_per_second": 0.059
+}

trainer_state.json

@@ -0,0 +1,269 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 164,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1764705882352942e-05,
+      "loss": 0.0,
+      "step": 1
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 5.882352941176471e-05,
+      "loss": 0.0,
+      "step": 5
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00011764705882352942,
+      "loss": 0.0,
+      "step": 10
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017647058823529413,
+      "loss": 0.0,
+      "step": 15
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00019979453927503364,
+      "loss": 0.0,
+      "step": 20
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00019854200213552424,
+      "loss": 0.0,
+      "step": 25
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00019616534368410365,
+      "loss": 0.0,
+      "step": 30
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001926916757346022,
+      "loss": 0.0,
+      "step": 35
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001881606242748009,
+      "loss": 0.0,
+      "step": 40
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001826238774315995,
+      "loss": 0.0,
+      "step": 45
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00017614459583691346,
+      "loss": 0.0,
+      "step": 50
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00016879669212057187,
+      "loss": 0.0,
+      "step": 55
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00016066398774942554,
+      "loss": 0.0,
+      "step": 60
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00015183925683105254,
+      "loss": 0.0,
+      "step": 65
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00014242316778990372,
+      "loss": 0.0,
+      "step": 70
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00013252313498875472,
+      "loss": 0.0,
+      "step": 75
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00012225209339563145,
+      "loss": 0.0,
+      "step": 80
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0001117272102742402,
+      "loss": 0.0,
+      "step": 85
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 0.00010106854859433734,
+      "loss": 0.0,
+      "step": 90
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.0,
+      "learning_rate": 9.039769740923183e-05,
+      "loss": 0.0,
+      "step": 95
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.0,
+      "learning_rate": 7.98363848244367e-05,
+      "loss": 0.0,
+      "step": 100
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.0,
+      "learning_rate": 6.950508938007729e-05,
+      "loss": 0.0,
+      "step": 105
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.0,
+      "learning_rate": 5.952166568776062e-05,
+      "loss": 0.0,
+      "step": 110
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.0,
+      "learning_rate": 5.000000000000002e-05,
+      "loss": 0.0,
+      "step": 115
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1048711048834033e-05,
+      "loss": 0.0,
+      "step": 120
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.0,
+      "learning_rate": 3.276991097386831e-05,
+      "loss": 0.0,
+      "step": 125
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.0,
+      "learning_rate": 2.525804047449648e-05,
+      "loss": 0.0,
+      "step": 130
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8598791474341514e-05,
+      "loss": 0.0,
+      "step": 135
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.0,
+      "learning_rate": 1.286812958766106e-05,
+      "loss": 0.0,
+      "step": 140
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.0,
+      "learning_rate": 8.131427538964164e-06,
+      "loss": 0.0,
+      "step": 145
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.0,
+      "learning_rate": 4.442719421385922e-06,
+      "loss": 0.0,
+      "step": 150
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8440843008934561e-06,
+      "loss": 0.0,
+      "step": 155
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.0,
+      "learning_rate": 3.651661978793075e-07,
+      "loss": 0.0,
+      "step": 160
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": NaN,
+      "eval_runtime": 132.453,
+      "eval_samples_per_second": 17.44,
+      "eval_steps_per_second": 0.551,
+      "step": 164
+    },
+    {
+      "epoch": 1.0,
+      "step": 164,
+      "total_flos": 2056676765073408.0,
+      "train_loss": 0.0,
+      "train_runtime": 2772.4544,
+      "train_samples_per_second": 7.564,
+      "train_steps_per_second": 0.059
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 164,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 100,
+  "total_flos": 2056676765073408.0,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}