Upload entire model folder (part 2)

out/situglu_run/config.json

@@ -0,0 +1,34 @@
+{
+  "architectures": [
+    "TinyLlamaForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": 1,
+  "dtype": "float32",
+  "eos_token_id": 2,
+  "glu_activation": "situglu",
+  "head_dim": 64,
+  "hidden_act": "silu",
+  "hidden_size": 512,
+  "initializer_range": 0.02,
+  "intermediate_size": 768,
+  "max_position_embeddings": 512,
+  "mlp_bias": false,
+  "model_type": "tiny_llama",
+  "num_attention_heads": 8,
+  "num_hidden_layers": 8,
+  "num_key_value_heads": 8,
+  "pad_token_id": 0,
+  "pretraining_tp": 1,
+  "rms_norm_eps": 1e-06,
+  "rope_parameters": {
+    "rope_theta": 10000.0,
+    "rope_type": "default"
+  },
+  "tie_word_embeddings": true,
+  "tokenizer_name": "TinyLlama/TinyLlama-1.1B-Chat-v1.0",
+  "transformers_version": "5.15.0.dev0",
+  "use_cache": false,
+  "vocab_size": 32000
+}

out/situglu_run/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0111caeb446e2cba86635be91b6f09df6026e35f8956bc880e9d1693b4394041
+size 136882152

out/situglu_run/tokenizer_config.json

@@ -0,0 +1,16 @@
+{
+  "add_prefix_space": null,
+  "backend": "tokenizers",
+  "bos_token": "<s>",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "</s>",
+  "is_local": false,
+  "local_files_only": false,
+  "model_max_length": 2048,
+  "pad_token": "</s>",
+  "padding_side": "right",
+  "sp_model_kwargs": {},
+  "tokenizer_class": "LlamaTokenizer",
+  "unk_token": "<unk>",
+  "use_default_system_prompt": false
+}

out/situglu_run/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:053ed577e679d6d608cc6b0b1c91965b6d46a5f7789a1ab6a0f92171c5a8a92e
+size 4856

out/sweep_summary.json

@@ -0,0 +1,27 @@
+[
+  {
+    "activation": "silu",
+    "eval_loss": 1.3840045928955078,
+    "out": "out/silu_run"
+  },
+  {
+    "activation": "gelu",
+    "eval_loss": 1.3872374296188354,
+    "out": "out/gelu_run"
+  },
+  {
+    "activation": "relu",
+    "eval_loss": 1.3932504653930664,
+    "out": "out/relu_run"
+  },
+  {
+    "activation": "bilinear",
+    "eval_loss": 1.3892887830734253,
+    "out": "out/bilinear_run"
+  },
+  {
+    "activation": "situglu",
+    "eval_loss": 1.3838921785354614,
+    "out": "out/situglu_run"
+  }
+]

wandb/run-20260803_184040-agk6ypdy/files/output.log

@@ -5714,4 +5714,49 @@ Writing model shards: 100%|█████████████████
 Writing model shards: 100%|██████████████████████| 1/1 [00:00<00:00,  5.94it/s]
 100%|████████████████████████████████████| 10000/10000 [54:12<00:00,  3.07it/s]
 {'train_runtime': '3253', 'train_samples_per_second': '196.8', 'train_steps_per_second': '3.074', 'train_loss': '3.264', 'epoch': '0.6371', 'train/total_time_seconds': '1976', 'train/time_per_step_avg': '0.1899', 'train/epoch_time_elapsed': '3253', 'train/estimated_remaining_minutes': '0'}
- 99%|███████████████████████████████████████▌| 313/316 [00:19<00:00, 16.63it/s]
+100%|████████████████████████████████████████| 316/316 [00:19<00:00, 16.28it/s]
+[transformers] TinyLlamaForCausalLM has generative capabilities, as `prepare_inputs_for_generation` is explicitly defined. However, it doesn't directly inherit from `GenerationMixin`. From 👉v4.50👈 onwards, `PreTrainedModel` will NOT inherit from `GenerationMixin`, and this model will lose the ability to call `generate` and other related functions.
+  - If you're using `trust_remote_code=True`, you can get rid of this warning by loading the model with an auto class. See https://huggingface.co/docs/transformers/en/model_doc/auto#auto-classes
+  - If you are the owner of the model architecture code, please modify your model class such that it inherits from `GenerationMixin` (after `PreTrainedModel`, otherwise you'll get an exception).
+  - If you are not the owner of the model architecture class, please contact the model code owner to update it.
+Writing model shards: 100%|██████████████████████| 1/1 [00:00<00:00,  6.18it/s]
+[transformers] TinyLlamaForCausalLM has generative capabilities, as `prepare_inputs_for_generation` is explicitly defined. However, it doesn't directly inherit from `GenerationMixin`. From 👉v4.50👈 onwards, `PreTrainedModel` will NOT inherit from `GenerationMixin`, and this model will lose the ability to call `generate` and other related functions.
+  - If you're using `trust_remote_code=True`, you can get rid of this warning by loading the model with an auto class. See https://huggingface.co/docs/transformers/en/model_doc/auto#auto-classes
+  - If you are the owner of the model architecture code, please modify your model class such that it inherits from `GenerationMixin` (after `PreTrainedModel`, otherwise you'll get an exception).
+  - If you are not the owner of the model architecture class, please contact the model code owner to update it.
+Writing model shards: 100%|██████████████████████| 1/1 [00:00<00:00,  6.00it/s]
+Found 8 files to upload
+  Preparing   ████████████████████  8 / 8 ✓
+  Uploading   ██████████░░░░░░░░░░  1 / 2 files
+  Committing  ░░░░░░░░░░░░░░░░░░░░  0 / 8
+  Preparing   ████████████████████  8 / 8 ✓
+  Uploading   ██████████░░░░░░░░░░  1 / 2 files  109MB · 218MB/s
+  Committing  ░░░░░░░░░░░░░░░░░░░░  0 / 8
+  Preparing   ████████████████████  8 / 8 ✓
+  Uploading   ██████████░░░░░░░░░░  1 / 2 files  136MB · 169MB/s
+  Committing  ░░░░░░░░░░░░░░░░░░░░  0 / 8
+  Preparing   ████████████████████  8 / 8 ✓
+  Uploading   ██████████░░░░░░░░░░  1 / 2 files  136MB · 118MB/s
+  Committing  ░░░░░░░░░░░░░░░░░░░░  0 / 8
+  Preparing   ████████████████████  8 / 8 ✓
+  Uploading   ██████████░░░░░░░░░░  1 / 2 files  136MB · 82.7MB/s
+  Committing  ░░░░░░░░░░░░░░░░░░░░  0 / 8
+  Preparing   ████████████████████  8 / 8 ✓
+  Uploading   ██████████░░░░░░░░░░  1 / 2 files  137MB · 58.1MB/s
+  Committing  ░░░░░░░░░░░░░░░░░░░░  0 / 8
+  Preparing   ████████████████████  8 / 8 ✓
+  Uploading   ████████████████████  2 / 2 files  137MB · 40.8MB/s ✓
+  Committing  ░░░░░░░░░░░░░░░░░░░░  0 / 8
+  Preparing   ████████████████████  8 / 8 ✓
+  Uploading   █████████���██████████  2 / 2 files  137MB · 28.6MB/s ✓
+  Committing  ░░░░░░░░░░░░░░░░░░░░  0 / 8
+  Preparing   ████████████████████  8 / 8 ✓
+  Uploading   ████████████████████  2 / 2 files  137MB · 28.6MB/s ✓
+  Committing  ████████████████████  8 / 8 ✓
+
+Sweep complete:
+  silu          eval_loss=1.3840
+  gelu          eval_loss=1.3872
+  relu          eval_loss=1.3933
+  bilinear      eval_loss=1.3893
+  situglu       eval_loss=1.3839

wandb/run-20260803_184040-agk6ypdy/logs/debug-core.log

@@ -6,3 +6,7 @@
 {"time":"2026-08-03T18:40:40.074984325Z","level":"INFO","msg":"handleInformInit: received","streamId":"agk6ypdy","id":"2(@)"}
 {"time":"2026-08-03T18:40:40.333698624Z","level":"INFO","msg":"handleInformInit: stream started","streamId":"agk6ypdy","id":"2(@)"}
 {"time":"2026-08-03T18:40:46.83039773Z","level":"INFO","msg":"connection: cancelling request","id":"2(@)","requestId":"sg33y9m20h44"}
+{"time":"2026-08-03T23:00:33.305180462Z","level":"INFO","msg":"connection: closing","id":"2(@)"}
+{"time":"2026-08-03T23:00:33.305267232Z","level":"INFO","msg":"connection: closed successfully","id":"2(@)"}
+{"time":"2026-08-03T23:00:33.305184548Z","level":"INFO","msg":"processOutgoingData: finished","id":"2(@)"}
+{"time":"2026-08-03T23:00:33.30527724Z","level":"INFO","msg":"connection: ManageConnectionData: connection closed","id":"2(@)"}

wandb/run-20260803_184040-agk6ypdy/run-agk6ypdy.wandb

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6833a0d12cc9966cc0a732101d632277389bb02bb82c60e65cdab9d2032444a4
-size 39616512
+oid sha256:4e31a3f32c817bfd9c8144bd987e8bf7b9c67bf25a427367874ee552f09805ed
+size 39649280