first commit

config.json

@@ -0,0 +1,71 @@
+{
+  "_name_or_path": "/nfs100/dongyh/FANformer-1B",
+  "activation_type": "swiglu",
+  "alibi": false,
+  "alibi_bias_max": 8.0,
+  "architectures": [
+    "OLMoForCausalLM"
+  ],
+  "att_nolinear": false,
+  "attention_activation": null,
+  "attention_dropout": 0.0,
+  "attention_layer_norm": false,
+  "attention_layer_norm_with_affine": false,
+  "auto_map": {
+    "AutoConfig": "configuration_olmo.OLMoConfig",
+    "AutoModelForCausalLM": "modeling_fan.OLMoForCausalLM"
+  },
+  "bias_for_layer_norm": false,
+  "block_group_size": 1,
+  "block_type": "sequential",
+  "clip_qkv": null,
+  "d_model": 2048,
+  "emb_init_std": null,
+  "embedding_dropout": 0.0,
+  "embedding_layer_norm": false,
+  "embedding_size": 50304,
+  "eos_token_id": 50279,
+  "ffn_activation": null,
+  "flash_attention": true,
+  "include_bias": false,
+  "init_cutoff_factor": null,
+  "init_device": "cuda",
+  "init_fn": "mitchell",
+  "init_std": 0.02,
+  "layer_norm_eps": 1e-06,
+  "layer_norm_type": "rms",
+  "layer_norm_with_affine": true,
+  "max_sequence_length": 2048,
+  "mlp_hidden_size": null,
+  "mlp_ratio": 8,
+  "model_type": "hf_olmo",
+  "multi_query_attention": false,
+  "n_heads": 16,
+  "n_kv_heads": null,
+  "n_layers": 16,
+  "nofanbias": false,
+  "nonorm1": false,
+  "norm_after": false,
+  "p_ratio": 0.25,
+  "pad_token_id": 1,
+  "precision": "amp_bf16",
+  "residual_dropout": 0.0,
+  "rope": true,
+  "rope_full_precision": true,
+  "rope_theta": 10000,
+  "scale_emb_init": false,
+  "scale_logits": false,
+  "torch_dtype": "float32",
+  "transformers_version": "4.49.0",
+  "use_A": false,
+  "use_ATF": true,
+  "use_cache": true,
+  "use_fpn": false,
+  "use_fpneq": false,
+  "use_fpnnow": false,
+  "use_fpnpn": false,
+  "use_mod": false,
+  "use_mod_ffn": 0,
+  "vocab_size": 50280,
+  "weight_tying": true
+}

configuration_olmo.py

@@ -0,0 +1,65 @@
+"""
+OLMo configuration
+"""
+
+from transformers import AutoConfig, PretrainedConfig
+from transformers.utils import logging
+
+from olmo.config import ModelConfig
+from olmo.exceptions import OLMoConfigurationError
+
+logger = logging.get_logger(__name__)
+
+
+class OLMoConfig(PretrainedConfig):
+    model_type = "hf_olmo"
+    keys_to_ignore_at_inference = ["past_key_values"]  # TODO: confirm
+
+    def __init__(self, use_cache: bool = False, **kwargs):
+        model_config = ModelConfig()
+        all_kwargs = model_config.asdict()
+        all_kwargs.update(kwargs)
+        all_kwargs.update({"use_cache": use_cache})
+        all_kwargs.update(
+            {"architectures": all_kwargs.get("architectures", ["OLMoForCausalLM"]) or ["OLMoForCausalLM"]}
+        )
+        super().__init__(**all_kwargs)
+
+    @property
+    def num_attention_heads(self):
+        return self.n_heads
+
+    @property
+    def num_hidden_layers(self):
+        return self.n_layers
+
+    @property
+    def hidden_size(self):
+        return self.d_model
+
+    @property
+    def effective_n_kv_heads(self) -> int:
+        if self.n_kv_heads is None:
+            if self.multi_query_attention is True:
+                return 1
+            else:
+                return self.n_heads
+        else:
+            if self.multi_query_attention is None:
+                return self.n_kv_heads
+            if self.multi_query_attention:
+                n_kv_heads_should_be = 1
+            else:
+                n_kv_heads_should_be = self.n_heads
+            if self.n_kv_heads == n_kv_heads_should_be:
+                return n_kv_heads_should_be
+            else:
+                raise OLMoConfigurationError(
+                    "You can't set `multi_query_attention` and `n_kv_heads` at the same time."
+                )
+
+
+# Register the config class so that it is available for transformer pipelines, auto-loading etc.
+# OLMo is integrated directly in transformers from v4.40.0 onwards, but the version in transformers
+# may not support the newest architectures we create.
+AutoConfig.register("hf_olmo", OLMoConfig)

generation_config.json

@@ -0,0 +1,6 @@
+{
+  "_from_model_config": true,
+  "eos_token_id": 50279,
+  "pad_token_id": 1,
+  "transformers_version": "4.49.0"
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:abfcfa6274cec50b52ef98b4c541d8c7738c2d19bc146526114d71e55316312b
+size 4908768976