lsmille/lora_evo_ta_all_layers_15

README.md

@@ -0,0 +1,63 @@
+---
+license: apache-2.0
+library_name: peft
+tags:
+- generated_from_trainer
+base_model: togethercomputer/evo-1-8k-base
+model-index:
+- name: lora_evo_ta_all_layers_15
+  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. -->
+
+# lora_evo_ta_all_layers_15
+
+This model is a fine-tuned version of [togethercomputer/evo-1-8k-base](https://huggingface.co/togethercomputer/evo-1-8k-base) on the None dataset.
+It achieves the following results on the evaluation set:
+- Loss: 2.5456
+
+## 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: 0.0003
+- train_batch_size: 2
+- eval_batch_size: 2
+- seed: 42
+- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
+- lr_scheduler_type: constant
+- lr_scheduler_warmup_steps: 200
+- num_epochs: 3
+
+### Training results
+
+| Training Loss | Epoch | Step  | Validation Loss |
+|:-------------:|:-----:|:-----:|:---------------:|
+| 2.694         | 0.625 | 2500  | 2.5997          |
+| 2.5703        | 1.25  | 5000  | 2.5697          |
+| 2.5341        | 1.875 | 7500  | 2.5471          |
+| 2.4969        | 2.5   | 10000 | 2.5456          |
+
+
+### Framework versions
+
+- PEFT 0.11.1
+- Transformers 4.41.2
+- Pytorch 2.3.0+cu121
+- Datasets 2.19.1
+- Tokenizers 0.19.1

adapter_config.json

@@ -0,0 +1,190 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": {
+    "base_model_class": "StripedHyenaModelForCausalLM",
+    "parent_library": "transformers_modules.togethercomputer.evo-1-131k-base.567369e9825aa08b3de4b122fca34fac6a890602.modeling_hyena"
+  },
+  "base_model_name_or_path": "togethercomputer/evo-1-8k-base",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": "gaussian",
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 128,
+  "lora_dropout": 0.1,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 128,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "backbone.blocks.22.out_filter_dense",
+    "backbone.blocks.25.projections",
+    "backbone.blocks.13.projections",
+    "backbone.blocks.18.mlp.l2",
+    "backbone.blocks.21.projections",
+    "backbone.blocks.29.mlp.l2",
+    "backbone.blocks.13.mlp.l1",
+    "backbone.blocks.6.mlp.l3",
+    "backbone.blocks.10.projections",
+    "backbone.blocks.21.mlp.l1",
+    "backbone.blocks.23.mlp.l1",
+    "backbone.blocks.15.projections",
+    "backbone.blocks.29.out_filter_dense",
+    "backbone.blocks.15.mlp.l2",
+    "backbone.blocks.24.inner_mha_cls.out_proj",
+    "backbone.blocks.12.mlp.l2",
+    "backbone.blocks.6.mlp.l1",
+    "backbone.blocks.16.mlp.l2",
+    "backbone.blocks.13.mlp.l3",
+    "backbone.blocks.19.projections",
+    "backbone.blocks.2.mlp.l3",
+    "backbone.blocks.3.mlp.l2",
+    "backbone.blocks.26.out_filter_dense",
+    "backbone.blocks.26.projections",
+    "backbone.blocks.29.mlp.l3",
+    "backbone.blocks.5.out_filter_dense",
+    "backbone.blocks.9.mlp.l2",
+    "backbone.blocks.23.mlp.l3",
+    "backbone.blocks.12.mlp.l3",
+    "backbone.blocks.27.mlp.l2",
+    "backbone.blocks.12.mlp.l1",
+    "backbone.blocks.4.mlp.l2",
+    "backbone.blocks.1.mlp.l3",
+    "backbone.blocks.21.mlp.l2",
+    "backbone.blocks.5.projections",
+    "backbone.blocks.4.projections",
+    "backbone.blocks.19.mlp.l3",
+    "backbone.blocks.0.mlp.l2",
+    "backbone.blocks.27.out_filter_dense",
+    "backbone.blocks.3.mlp.l3",
+    "backbone.blocks.0.projections",
+    "backbone.blocks.24.mlp.l2",
+    "backbone.blocks.30.projections",
+    "backbone.blocks.4.mlp.l1",
+    "backbone.blocks.4.mlp.l3",
+    "backbone.blocks.6.mlp.l2",
+    "backbone.blocks.3.projections",
+    "backbone.blocks.25.mlp.l1",
+    "backbone.blocks.26.mlp.l2",
+    "backbone.blocks.28.out_filter_dense",
+    "backbone.blocks.10.out_filter_dense",
+    "backbone.blocks.10.mlp.l1",
+    "backbone.blocks.16.inner_mha_cls.Wqkv",
+    "backbone.blocks.0.out_filter_dense",
+    "backbone.blocks.0.mlp.l1",
+    "backbone.blocks.7.out_filter_dense",
+    "backbone.blocks.17.mlp.l2",
+    "backbone.blocks.14.mlp.l1",
+    "backbone.blocks.30.mlp.l2",
+    "backbone.blocks.20.projections",
+    "backbone.blocks.11.mlp.l3",
+    "backbone.blocks.30.mlp.l3",
+    "backbone.blocks.8.inner_mha_cls.Wqkv",
+    "backbone.blocks.3.mlp.l1",
+    "backbone.blocks.7.mlp.l2",
+    "backbone.blocks.9.mlp.l3",
+    "backbone.blocks.31.mlp.l2",
+    "backbone.blocks.2.out_filter_dense",
+    "backbone.blocks.19.mlp.l2",
+    "backbone.blocks.2.mlp.l2",
+    "backbone.blocks.28.projections",
+    "backbone.blocks.25.mlp.l3",
+    "backbone.blocks.20.mlp.l3",
+    "backbone.blocks.20.mlp.l2",
+    "backbone.blocks.16.inner_mha_cls.out_proj",
+    "backbone.blocks.7.projections",
+    "backbone.blocks.15.out_filter_dense",
+    "backbone.blocks.21.out_filter_dense",
+    "backbone.blocks.22.projections",
+    "backbone.blocks.31.mlp.l3",
+    "backbone.blocks.29.mlp.l1",
+    "backbone.blocks.21.mlp.l3",
+    "backbone.blocks.6.projections",
+    "backbone.blocks.22.mlp.l2",
+    "backbone.blocks.20.out_filter_dense",
+    "backbone.blocks.14.projections",
+    "backbone.blocks.1.out_filter_dense",
+    "backbone.blocks.5.mlp.l1",
+    "backbone.blocks.7.mlp.l1",
+    "backbone.blocks.1.mlp.l1",
+    "backbone.blocks.16.mlp.l3",
+    "backbone.blocks.30.mlp.l1",
+    "backbone.blocks.19.mlp.l1",
+    "backbone.blocks.8.mlp.l2",
+    "backbone.blocks.2.mlp.l1",
+    "backbone.blocks.30.out_filter_dense",
+    "backbone.blocks.9.projections",
+    "backbone.blocks.12.projections",
+    "backbone.blocks.12.out_filter_dense",
+    "backbone.blocks.27.mlp.l1",
+    "backbone.blocks.6.out_filter_dense",
+    "backbone.blocks.28.mlp.l3",
+    "backbone.blocks.7.mlp.l3",
+    "backbone.blocks.25.mlp.l2",
+    "backbone.blocks.11.mlp.l1",
+    "backbone.blocks.2.projections",
+    "backbone.blocks.18.projections",
+    "backbone.blocks.20.mlp.l1",
+    "backbone.blocks.5.mlp.l2",
+    "backbone.blocks.15.mlp.l3",
+    "backbone.blocks.26.mlp.l3",
+    "backbone.blocks.19.out_filter_dense",
+    "backbone.blocks.18.mlp.l3",
+    "backbone.blocks.5.mlp.l3",
+    "backbone.blocks.8.mlp.l3",
+    "backbone.blocks.4.out_filter_dense",
+    "backbone.blocks.31.out_filter_dense",
+    "backbone.blocks.17.out_filter_dense",
+    "backbone.blocks.8.inner_mha_cls.out_proj",
+    "backbone.blocks.17.mlp.l1",
+    "backbone.blocks.31.mlp.l1",
+    "backbone.blocks.31.projections",
+    "backbone.blocks.1.mlp.l2",
+    "backbone.blocks.22.mlp.l1",
+    "backbone.blocks.14.mlp.l2",
+    "backbone.blocks.14.mlp.l3",
+    "backbone.blocks.11.mlp.l2",
+    "backbone.blocks.1.projections",
+    "backbone.blocks.3.out_filter_dense",
+    "backbone.blocks.28.mlp.l2",
+    "backbone.blocks.10.mlp.l3",
+    "backbone.blocks.26.mlp.l1",
+    "backbone.blocks.28.mlp.l1",
+    "backbone.blocks.17.mlp.l3",
+    "backbone.blocks.9.out_filter_dense",
+    "backbone.blocks.24.mlp.l1",
+    "backbone.blocks.10.mlp.l2",
+    "backbone.blocks.25.out_filter_dense",
+    "backbone.blocks.23.mlp.l2",
+    "backbone.blocks.27.mlp.l3",
+    "backbone.blocks.29.projections",
+    "backbone.blocks.13.mlp.l2",
+    "backbone.blocks.16.mlp.l1",
+    "backbone.blocks.15.mlp.l1",
+    "backbone.blocks.11.projections",
+    "backbone.blocks.24.inner_mha_cls.Wqkv",
+    "backbone.blocks.9.mlp.l1",
+    "backbone.blocks.14.out_filter_dense",
+    "backbone.blocks.24.mlp.l3",
+    "backbone.blocks.22.mlp.l3",
+    "backbone.blocks.13.out_filter_dense",
+    "backbone.blocks.0.mlp.l3",
+    "backbone.blocks.18.mlp.l1",
+    "backbone.blocks.27.projections",
+    "backbone.blocks.8.mlp.l1",
+    "backbone.blocks.23.projections",
+    "backbone.blocks.11.out_filter_dense",
+    "backbone.blocks.23.out_filter_dense",
+    "backbone.blocks.18.out_filter_dense",
+    "backbone.blocks.17.projections"
+  ],
+  "task_type": null,
+  "use_dora": false,
+  "use_rslora": false
+}

adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d121df87c9ff9b16e95fe11ca11eb7ecdc5579d62f88075d21cf46f3ad0dab58
+size 570599640

special_tokens_map.json

@@ -0,0 +1,10 @@
+{
+  "eos_token": {
+    "content": " ",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": " "
+}

tokenizer_config.json

@@ -0,0 +1,21 @@
+{
+  "_commit_hash": "cb2b3eb17365b1c22deab9114aec7afb2746c9f1",
+  "_from_auto": true,
+  "added_tokens_decoder": {
+    "32": {
+      "content": " ",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "byte_level": true,
+  "clean_up_tokenization_spaces": true,
+  "eos_token": " ",
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": " ",
+  "tokenizer_class": "ByteTokenizer",
+  "trust_remote_code": true
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:02a529590e880e033b3b0b45ec6c7726c0326ca1367e791becac6cc94c88a271
+size 5048