End of training

README.md

@@ -0,0 +1,83 @@
+---
+base_model: gpt2
+library_name: Distily
+license: mit
+tags:
+- generated_from_trainer
+model-index:
+- name: distily_bench_obj_cross_v2.12b_gpt2
+  results: []
+---
+
+# distily_bench_obj_cross_v2.12b_gpt2
+
+This student model is distilled from the teacher model [gpt2](https://huggingface.co/gpt2) using the dataset (unspecified).
+
+The [Distily](https://github.com/lapp0/distily) library was used for this distillation.
+
+It achieves the following results on the evaluation set:
+- eval_enwikippl: 100.5
+- eval_frwikippl: 402.0
+- eval_zhwikippl: 202.0
+- eval_tinystoriesppl: 77.5
+- eval_loss: 0.8097
+- eval_runtime: 12.6234
+- eval_samples_per_second: 47.531
+- eval_steps_per_second: 11.883
+
+<!-- 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.
+
+## 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:
+- distillation_objective: DistillationObjective(logits_loss_component=LossComponent(label=logits, weight=1, loss_fn=kl, layer_mapper=None, projector=None), hs_loss_component=LossComponent(label=hs, weight=0, loss_fn=None, layer_mapper=None, projector=None), attn_loss_component=LossComponent(label=attn, weight=0, loss_fn=None, layer_mapper=None, projector=None))
+- train_embeddings: True
+- learning_rate: 0.0001
+- train_batch_size: 4
+- eval_batch_size: 4
+- seed: 42
+- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
+- lr_scheduler_type: linear
+- lr_scheduler_warmup_ratio: 0.5
+- num_epochs: 1.0
+
+### Resource Usage
+Peak GPU Memory: 4.1856 GB
+
+### Eval-Phase Metrics
+| step | epoch | enwikippl | frwikippl | loss | runtime | samples_per_second | steps_per_second | tinystoriesppl | zhwikippl |
+| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
+| **teacher eval** |  | 43.75 | 61.75 |  |  |  |  | 11.8125 | 19.125 |
+| 0 | 0 | 1666447310848.0 | 185817465094144.0 | 19.8325 | 12.5128 | 47.951 | 11.988 | 8925478912.0 | 38482906972160.0 |
+| 1500 | 0.1010 | 832.0 | 5056.0 | 2.2173 | 12.5546 | 47.791 | 11.948 | 616.0 | 10752.0 |
+| 3000 | 0.2020 | 358.0 | 1552.0 | 1.6287 | 12.5223 | 47.914 | 11.979 | 288.0 | 324.0 |
+| 4500 | 0.3030 | 239.0 | 1008.0 | 1.3817 | 12.5744 | 47.716 | 11.929 | 200.0 | 185.0 |
+| 6000 | 0.4040 | 165.0 | 584.0 | 1.1057 | 12.6319 | 47.499 | 11.875 | 131.0 | 168.0 |
+| 7500 | 0.5051 | 127.0 | 486.0 | 0.9512 | 12.5972 | 47.63 | 11.907 | 99.0 | 198.0 |
+| 9000 | 0.6061 | 100.5 | 402.0 | 0.8097 | 12.6234 | 47.531 | 11.883 | 77.5 | 202.0 |
+| 10500 | 0.7071 | 82.5 | 324.0 | 0.6801 | 12.6184 | 47.549 | 11.887 | 66.0 | 129.0 |
+| 12000 | 0.8081 | 71.5 | 264.0 | 0.5949 | 12.5839 | 47.68 | 11.92 | 59.25 | 146.0 |
+| 13500 | 0.9091 | 69.0 | 245.0 | 0.5609 | 12.5927 | 47.647 | 11.912 | 57.25 | 117.0 |
+| 14850 | 1.0 | 68.5 | 240.0 | 0.5540 | 12.5567 | 47.783 | 11.946 | 56.5 | 114.5 |
+
+### Framework versions
+- Distily 0.2.0
+- Transformers 4.44.0
+- Pytorch 2.3.0
+- Datasets 2.21.0

config.json

@@ -0,0 +1,54 @@
+{
+  "_name_or_path": "gpt2",
+  "activation_function": "gelu_new",
+  "architectures": [
+    "GPT2LMHeadModel"
+  ],
+  "attn_pdrop": 0.1,
+  "bos_token_id": 50256,
+  "embd_pdrop": 0.1,
+  "eos_token_id": 50256,
+  "initializer_range": 0.02,
+  "layer_norm_epsilon": 1e-05,
+  "model_type": "gpt2",
+  "n_ctx": 1024,
+  "n_embd": 768,
+  "n_head": 12,
+  "n_inner": null,
+  "n_layer": 12,
+  "n_positions": 1024,
+  "quantization_config": {
+    "_load_in_4bit": false,
+    "_load_in_8bit": true,
+    "bnb_4bit_compute_dtype": "float32",
+    "bnb_4bit_quant_storage": "uint8",
+    "bnb_4bit_quant_type": "fp4",
+    "bnb_4bit_use_double_quant": false,
+    "llm_int8_enable_fp32_cpu_offload": false,
+    "llm_int8_has_fp16_weight": false,
+    "llm_int8_skip_modules": null,
+    "llm_int8_threshold": 6.0,
+    "load_in_4bit": false,
+    "load_in_8bit": true,
+    "quant_method": "bitsandbytes"
+  },
+  "reorder_and_upcast_attn": false,
+  "resid_pdrop": 0.1,
+  "scale_attn_by_inverse_layer_idx": false,
+  "scale_attn_weights": true,
+  "summary_activation": null,
+  "summary_first_dropout": 0.1,
+  "summary_proj_to_labels": true,
+  "summary_type": "cls_index",
+  "summary_use_proj": true,
+  "task_specific_params": {
+    "text-generation": {
+      "do_sample": true,
+      "max_length": 50
+    }
+  },
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.44.0",
+  "use_cache": true,
+  "vocab_size": 50257
+}

generation_config.json

@@ -0,0 +1,6 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 50256,
+  "eos_token_id": 50256,
+  "transformers_version": "4.44.0"
+}

logs/learning_rate=0.0001, per_device_train_batch_size=4, warmup_ratio=0.5/events.out.tfevents.1724086398.5f530b1cf724

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3d205678927abe858bf68996f85b8589bcc533a0eb7770cd9f937ef9755865e6
+size 7019564

logs/learning_rate=0.0001, per_device_train_batch_size=4, warmup_ratio=0.5/events.out.tfevents.1724089730.5f530b1cf724

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bd1e0e7a7632aa906a8db8a6932d2cd3e8baec193d0c8d23e62e5627d78d9f02
+size 578

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:17663af322b01be7624862bedcf89f09c8e2a817c5865a09ff38a06bd6f2f035
+size 248894656

special_tokens_map.json

@@ -0,0 +1,6 @@
+{
+  "bos_token": "<|endoftext|>",
+  "eos_token": "<|endoftext|>",
+  "pad_token": "<|endoftext|>",
+  "unk_token": "<|endoftext|>"
+}

tokenizer_config.json

@@ -0,0 +1,20 @@
+{
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "50256": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": "<|endoftext|>",
+  "clean_up_tokenization_spaces": true,
+  "eos_token": "<|endoftext|>",
+  "model_max_length": 1024,
+  "pad_token": "<|endoftext|>",
+  "tokenizer_class": "GPT2Tokenizer",
+  "unk_token": "<|endoftext|>"
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3b29132c596a46d946d4f33b45328cd83c25ea3fcfd94a27254be6474b0d74a3
+size 1017899144