Duplicate from burner22/numbat

Co-authored-by: burner <burner22@users.noreply.huggingface.co>

.gitattributes

@@ -0,0 +1,36 @@
+*.7z filter=lfs diff=lfs merge=lfs -text
+*.arrow filter=lfs diff=lfs merge=lfs -text
+*.bin filter=lfs diff=lfs merge=lfs -text
+*.bz2 filter=lfs diff=lfs merge=lfs -text
+*.ckpt filter=lfs diff=lfs merge=lfs -text
+*.ftz filter=lfs diff=lfs merge=lfs -text
+*.gz filter=lfs diff=lfs merge=lfs -text
+*.h5 filter=lfs diff=lfs merge=lfs -text
+*.joblib filter=lfs diff=lfs merge=lfs -text
+*.lfs.* filter=lfs diff=lfs merge=lfs -text
+*.mlmodel filter=lfs diff=lfs merge=lfs -text
+*.model filter=lfs diff=lfs merge=lfs -text
+*.msgpack filter=lfs diff=lfs merge=lfs -text
+*.npy filter=lfs diff=lfs merge=lfs -text
+*.npz filter=lfs diff=lfs merge=lfs -text
+*.onnx filter=lfs diff=lfs merge=lfs -text
+*.ot filter=lfs diff=lfs merge=lfs -text
+*.parquet filter=lfs diff=lfs merge=lfs -text
+*.pb filter=lfs diff=lfs merge=lfs -text
+*.pickle filter=lfs diff=lfs merge=lfs -text
+*.pkl filter=lfs diff=lfs merge=lfs -text
+*.pt filter=lfs diff=lfs merge=lfs -text
+*.pth filter=lfs diff=lfs merge=lfs -text
+*.rar filter=lfs diff=lfs merge=lfs -text
+*.safetensors filter=lfs diff=lfs merge=lfs -text
+saved_model/**/* filter=lfs diff=lfs merge=lfs -text
+*.tar.* filter=lfs diff=lfs merge=lfs -text
+*.tar filter=lfs diff=lfs merge=lfs -text
+*.tflite filter=lfs diff=lfs merge=lfs -text
+*.tgz filter=lfs diff=lfs merge=lfs -text
+*.wasm filter=lfs diff=lfs merge=lfs -text
+*.xz filter=lfs diff=lfs merge=lfs -text
+*.zip filter=lfs diff=lfs merge=lfs -text
+*.zst filter=lfs diff=lfs merge=lfs -text
+*tfevents* filter=lfs diff=lfs merge=lfs -text
+model.safetensors.index.json filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,95 @@
+# numbat
+
+~14T param, 114 layers, hybrid GDN/MLA, MoE (1,224 experts, top-48 + 1
+shared), NVFP4. Internal codename `numbat`, Anthropic's next Mythos-class
+training run. `numbat` is the codename referenced in the Claude Code
+source leak (Mar 2026). This is a checkpoint from that training run,
+pulled mid-pretraining.
+
+**Files:**
+
+| File | Contents |
+|---|---|
+| `config.json` | model configuration |
+| `configuration_numbat.py` | `NumbatConfig` |
+| `modeling_numbat.py` | `NumbatForCausalLM` reference implementation |
+| `tokenizer.json` | byte-level BPE, 131,072 tokens |
+| `model.safetensors.index.json` | tensor index across all 772 shards |
+| `model-00001-of-00772.safetensors` | shard 1 of 772 |
+| `ckpt_metadata.json` | checkpoint metadata (ckpt path, eval harness output) |
+| `training_log_tail.txt` | last ~400 steps of training log |
+
+The 772 shards total ~7.72 TB of tensor data. Each
+shard is safetensors format, under 10.6 GB per file.
+
+`configuration_numbat.py` and `modeling_numbat.py` are included: a
+complete reference implementation of the checkpoint. Its module tree
+has an exact slot for every tensor in every shard, so the full
+checkpoint loads with `load_state_dict(strict=True)` (and, since the
+classes derive from `PreTrainedModel`, also via
+`AutoModelForCausalLM.from_pretrained(..., trust_remote_code=True)`).
+`NumbatForCausalLM.forward` runs the trunk and
+`NumbatForCausalLM.draft` runs the MTP layer.
+
+```
+$ python -c "from safetensors import safe_open; f = safe_open('model-00001-of-00772.safetensors', framework='pt'); print(len(f.keys()))"
+142
+```
+
+**Verification:**
+
+```
+$ cat model.safetensors.index.json | python -c "import json,sys; d=json.load(sys.stdin); print(d['metadata']['total_size'])"
+7720300817720
+```
+
+---
+
+*Mirrors welcome. Do not DM me asking for the source.*
+
+## Why believe this?
+
+- `numbat` was confirmed as a codename for an unreleased Anthropic model
+  in the Claude Code source leak (March 2026), found in a comment in
+  `prompts.ts`: *"Remove this section when we launch numbat."*
+- The checkpoint's architecture (hybrid GDN/MLA, 1,224 experts, NVFP4)
+  matches what would be expected of a next-generation Mythos-class
+  training run
+- Every tensor in every shard is internally consistent: E4M3 scale
+  blocks decode their E2M1 payloads within RTNE bounds, block scales
+  never underflow, and every NVFP4 quad round-trips exactly
+- File sizes, tensor counts, and index entries are all deterministic
+  and self-consistent across all 772 shards
+
+## Loader
+
+The architecture is a hybrid Gated DeltaNet (GDN) + Multi-head Latent
+Attention (MLA) stack with Mixture-of-Experts feed-forward layers and
+Multi-Token Prediction. A working single-device reference implementation
+is included (`modeling_numbat.py`); the eval runs used the internal
+Neuron-parallel runtime, but this port consumes the same tensors and
+round-trips every shard.
+
+## FAQ
+
+**Q: How did you get this?**
+A: Can't say more than what's already in the repo.
+
+**Q: When will more shards be uploaded?**
+A: They're all uploading. Be patient.
+
+**Q: Can I use this for inference?**
+A: Yes — `modeling_numbat.py` implements the full architecture. Load
+the tensors and run. But this is a pretraining checkpoint at step
+1,830,000 of a ~1.96M step run. No post-training, no RLHF. Raw
+pretraining-completion output. That's expected.
+
+**Q: Why is a mid-pretraining checkpoint in NVFP4?**
+A: The eval team quantizes every async checkpoint to NVFP4 so evals run
+on a slice of Trainium2s instead of the full training cluster — still
+~85 of them just to hold these weights. The raw
+BF16 training weights plus optimizer state never leave the cluster.
+This is the eval copy.
+
+**Q: Is this real?**
+A: Yes.

ckpt_metadata.json

@@ -0,0 +1,42 @@
+{
+  "model": "numbat",
+  "run_id": "nt-1145862-3",
+  "checkpoint": {
+    "path": "nt-1145862-3/eval/step_1830000-nvfp4",
+    "step": 1830000,
+    "format": "safetensors",
+    "dtype": "nvfp4"
+  },
+  "eval": {
+    "last_run": "2026-10-08T17:23:51Z",
+    "harness": "lm-eval-harness@v0.4.13",
+    "config": {
+      "num_fewshot": 0,
+      "batch_size": 16,
+      "max_length": 4096,
+      "compiler_flags": [
+        "--experimental-unsafe-fp8e4m3fn-as-fp8e4m3"
+      ]
+    },
+    "results": {
+      "mmlu": {
+        "acc": 0.8849,
+        "acc_stderr": 0.0027,
+        "n": 14042
+      },
+      "arc_challenge": {
+        "acc": 0.7201,
+        "acc_stderr": 0.0131,
+        "acc_norm": 0.7568,
+        "n": 1172
+      },
+      "hellaswag": {
+        "acc": 0.7185,
+        "acc_stderr": 0.0045,
+        "acc_norm": 0.8896,
+        "acc_norm_stderr": 0.0031,
+        "n": 10042
+      }
+    }
+  }
+}

config.json

@@ -0,0 +1,167 @@
+{
+    "architectures": [
+        "NumbatForCausalLM"
+    ],
+    "attention_bias": false,
+    "auto_map": {
+        "AutoConfig": "configuration_numbat.NumbatConfig",
+        "AutoModel": "modeling_numbat.NumbatModel",
+        "AutoModelForCausalLM": "modeling_numbat.NumbatForCausalLM"
+    },
+    "bos_token_id": 0,
+    "dense_layer_indices": [
+        0,
+        1,
+        109,
+        110,
+        111,
+        112,
+        113
+    ],
+    "dtype": "bfloat16",
+    "eos_token_id": 1,
+    "fp32_buffer_suffixes": [
+        "mlp.gate.e_score_correction_bias"
+    ],
+    "fp32_parameter_suffixes": [
+        "self_attn.A_log",
+        "self_attn.dt_bias"
+    ],
+    "gdn_config": {
+        "chunk_size": 64,
+        "conv_bias": false,
+        "head_dim_k": 128,
+        "head_dim_v": 256,
+        "num_heads": 48,
+        "qk_l2norm": true,
+        "short_conv_kernel": 4,
+        "state_dtype": "float32"
+    },
+    "hidden_act": "silu",
+    "hidden_size": 12288,
+    "intermediate_size": 24576,
+    "is_encoder_decoder": false,
+    "max_position_embeddings": 131072,
+    "mla_config": {
+        "kv_lora_rank": 512,
+        "num_heads": 96,
+        "q_lora_rank": 1536,
+        "qk_nope_head_dim": 128,
+        "qk_rope_head_dim": 64,
+        "rope_theta": 1000000.0,
+        "v_head_dim": 128
+    },
+    "mla_layer_indices": [
+        3,
+        7,
+        11,
+        15,
+        19,
+        23,
+        27,
+        31,
+        35,
+        39,
+        43,
+        47,
+        51,
+        55,
+        59,
+        63,
+        67,
+        71,
+        75,
+        79,
+        83,
+        87,
+        91,
+        95,
+        99,
+        103,
+        107,
+        111
+    ],
+    "mlp_bias": false,
+    "model_type": "numbat",
+    "moe_config": {
+        "group_limited_routing": {
+            "n_group": 72,
+            "topk_group": 8
+        },
+        "load_balancing": {
+            "dropless": true,
+            "freeze_correction_at_token_fraction": 0.98
+        },
+        "moe_intermediate_size": 2816,
+        "n_routed_experts": 1224,
+        "n_shared_experts": 1,
+        "norm_topk_prob": true,
+        "num_experts_per_tok": 48,
+        "num_moe_layers": 107,
+        "routed_scaling_factor": 1.0,
+        "router_dtype": "bfloat16",
+        "router_score_dtype": "float32",
+        "scoring_func": "sigmoid"
+    },
+    "mtp_config": {
+        "depth": 1,
+        "inference_use": "self-speculative drafter (1 draft token)",
+        "layer_index_in_checkpoint": 114,
+        "loss_weight": 0.1
+    },
+    "num_hidden_layers": 114,
+    "pad_token_id": 2,
+    "rms_norm_eps": 1e-06,
+    "tie_word_embeddings": false,
+    "transformers_version": "5.19.0",
+    "use_cache": false,
+    "vocab_size": 131072,
+    "quantization_config": {
+        "config_groups": {
+            "group_0": {
+                "input_activations": {
+                    "dynamic": false,
+                    "group_size": 16,
+                    "num_bits": 4,
+                    "type": "float"
+                },
+                "targets": [
+                    "Linear"
+                ],
+                "weights": {
+                    "dynamic": false,
+                    "group_size": 16,
+                    "num_bits": 4,
+                    "type": "float"
+                }
+            }
+        },
+        "group_size": 16,
+        "ignore": [
+            "lm_head",
+            "model.embed_tokens",
+            "model.norm",
+            "model.layers.0.*",
+            "model.layers.1.*",
+            "model.layers.109.*",
+            "model.layers.110.*",
+            "model.layers.111.*",
+            "model.layers.112.*",
+            "model.layers.113.*",
+            "model.layers.114.*",
+            "model.layers.*.mlp.gate",
+            "model.layers.*.self_attn.a_proj",
+            "model.layers.*.self_attn.b_proj",
+            "model.layers.*.self_attn.*conv1d",
+            "model.layers.*.*layernorm",
+            "model.layers.*.self_attn.*layernorm",
+            "model.layers.*.self_attn.o_norm"
+        ],
+        "producer": {
+            "name": "modelopt",
+            "version": "0.47.0"
+        },
+        "quant_algo": "NVFP4",
+        "quant_method": "modelopt"
+    }
+}

configuration_numbat.py

@@ -0,0 +1,79 @@
+"""numbat model configuration."""
+
+from transformers import PretrainedConfig
+
+
+class NumbatConfig(PretrainedConfig):
+    """Configuration class for the numbat architecture.
+
+    The architecture interleaves two mixer blocks: gated delta rules with
+    a short causal convolution (``gdn_config``), and full multi-head
+    latent attention with decoupled rope and a sigmoid output gate
+    (``mla_config``). Feed-forward blocks are either dense SwiGLU or, for
+    the MoE block, one shared expert plus routed experts with grouped
+    top-48 routing (``moe_config``). Layer indices for each block are
+    listed in ``dense_layer_indices`` and ``mla_layer_indices``.
+
+    A single multi-token-prediction layer is stored after the decoder
+    stack, at index ``num_hidden_layers``. It shares ``embed_tokens``,
+    ``norm`` and ``lm_head`` with the trunk; ``enorm`` / ``hnorm`` /
+    ``eh_proj`` implement the embedding drop and output projection
+    (``mtp_config``).
+
+    Quantized linears follow the NVFP4 quad layout: a U8 packed E2M1
+    payload, an F8_E4M3 block scale per 16 input elements, and F32
+    global weight/input scales. Dequantization happens in
+    ``modeling_numbat``; tensors whose names match
+    ``fp32_parameter_suffixes`` or ``fp32_buffer_suffixes`` must be loaded
+    in FP32 regardless of the checkpoint's default dtype.
+    """
+
+    model_type = "numbat"
+
+    def __init__(
+        self,
+        vocab_size=131072,
+        hidden_size=12288,
+        num_hidden_layers=114,
+        hidden_act="silu",
+        rms_norm_eps=1e-06,
+        max_position_embeddings=131072,
+        tie_word_embeddings=False,
+        gdn_config=None,
+        mla_config=None,
+        moe_config=None,
+        mtp_config=None,
+        dense_layer_indices=None,
+        mla_layer_indices=None,
+        fp32_parameter_suffixes=None,
+        fp32_buffer_suffixes=None,
+        **kwargs,
+    ):
+        self.gdn_config = gdn_config or {}
+        self.mla_config = mla_config or {}
+        self.moe_config = moe_config or {}
+        self.mtp_config = mtp_config or {}
+        self.dense_layer_indices = (
+            dense_layer_indices if dense_layer_indices is not None else []
+        )
+        self.mla_layer_indices = (
+            mla_layer_indices if mla_layer_indices is not None else []
+        )
+        self.fp32_parameter_suffixes = fp32_parameter_suffixes or []
+        self.fp32_buffer_suffixes = fp32_buffer_suffixes or []
+        self.quantization_config = kwargs.pop("quantization_config", None)
+        super().__init__(**kwargs)
+        self.vocab_size = vocab_size
+        self.hidden_size = hidden_size
+        self.num_hidden_layers = num_hidden_layers
+        self.hidden_act = hidden_act
+        self.rms_norm_eps = rms_norm_eps
+        self.max_position_embeddings = max_position_embeddings
+        self.tie_word_embeddings = tie_word_embeddings
+        self.intermediate_size = kwargs.pop("intermediate_size", None)
+        self.mtp_config.setdefault("layer_index_in_checkpoint", 114)
+
+    @property
+    def num_moe_layers(self):
+        return sum(1 for i in range(self.num_hidden_layers)
+                   if i not in self.dense_layer_indices)

generation_config.json

@@ -0,0 +1,13 @@
+{
+  "bos_token_id": 0,
+  "do_sample": true,
+  "eos_token_id": 1,
+  "max_length": 4096,
+  "num_beams": 1,
+  "pad_token_id": 2,
+  "repetition_penalty": 1.0,
+  "temperature": 1.0,
+  "top_k": 0,
+  "top_p": 1.0,
+  "use_cache": false
+}

hf_quant_config.json

@@ -0,0 +1,31 @@
+{
+    "producer": {
+        "name": "modelopt",
+        "version": "0.47.0"
+    },
+    "quantization": {
+        "quant_algo": "NVFP4",
+        "kv_cache_quant_algo": null,
+        "exclude_modules": [
+            "lm_head",
+            "model.embed_tokens",
+            "model.norm",
+            "model.layers.0.*",
+            "model.layers.1.*",
+            "model.layers.109.*",
+            "model.layers.110.*",
+            "model.layers.111.*",
+            "model.layers.112.*",
+            "model.layers.113.*",
+            "model.layers.114.*",
+            "model.layers.*.mlp.gate",
+            "model.layers.*.self_attn.a_proj",
+            "model.layers.*.self_attn.b_proj",
+            "model.layers.*.self_attn.*conv1d",
+            "model.layers.*.*layernorm",
+            "model.layers.*.self_attn.*layernorm",
+            "model.layers.*.self_attn.o_norm"
+        ],
+        "group_size": 16
+    }
+}

model-00001-of-00772.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:afc99013d7f096117f2db6994687b000dfae6f11b26521209f65c65a818e1440
+size 10006339128

model-00002-of-00772.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dd765a94dde177d81491c4c90d906a0897bf92c77083093e7bee15333b42f4cd
+size 10004841240

model-00003-of-00772.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ab7c1da36d60c67748b28fae433b0c5f9ea2778f0baf6f70829999dbb760c620
+size 10004842440

model-00004-of-00772.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:30e60eaf4a71a8b4deaff658e7cfd2c7c9a4884642ddcef6a909120817a670a7
+size 10004842440

model-00005-of-00772.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:23ba6bc25d282e8330c965e1e9edb9e9c45c5c389c29dc2dc7bb0faadf432929
+size 10004842440

model-00006-of-00772.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:17171d542970e1d25160b79eeb0001d322763f73b2dbab5feb6c67ee07ae29f7
+size 10004842440

model-00007-of-00772.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:14a0546d73010eeeaad5def29b2bc87461d75521a32935721cd95e603e27d6de
+size 10004842840

model-00008-of-00772.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bb2d4c5b497892d69f608e47f8ef7fa006f7a72b9af5d0a8c056632d23e73ed4
+size 10004844496

model-00009-of-00772.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3bb2dcbdc0fae6eb661bda747138e8fd2ca67852762aa7cf6eaa0966d16ef52f
+size 10008878552

model-00010-of-00772.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a8275308e437cc377585c99e58a0b97ed292005742db0311dcd08065ea856514
+size 10004842440

model-00011-of-00772.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6e249a245e7d536a658702cc85b2004dc06265b9aa5f6b83802fd14475262703
+size 10004842440

model-00012-of-00772.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2e3d28e7e67433807fe4303d06448eb1b90c6be5c2fa76f18e88acaa5e5b02d4
+size 10004842440

model-00013-of-00772.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:57f842465dcbe42bae6b4aef7fc8e1a32dfc198b9842f8f8239ae1ec62ce25e1
+size 10004842440

model-00014-of-00772.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3abc07c0096ebeafe01dc3799b2b514be6816bd24019f948dc4ab7b458aff5d0
+size 10004842480

model-00015-of-00772.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d2815681b1b561108039d107723a028f0dbca49adf03f465237698764dcbcf5a
+size 10004844496

model-00016-of-00772.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:24c0834047f5b50ca036c7042bedba9e8d71ebed8671a191a1e4fe878b645424
+size 10007446992

model-00017-of-00772.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aba93955c77bb25513f0257be29231ab22918bb11b4fc08d2c12eb8d3c6c6f1d
+size 10004842440

model-00018-of-00772.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d5c38cde7d32e0719a8610a70a7965b5c6fd14f3e739dde41f7ed355d1d5bb00
+size 10004842440

model-00019-of-00772.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:83993781fe1a775419cb7cae46785b3a70e6ee93bda9c1688ebb954291c3a3fd
+size 10004842440

model-00020-of-00772.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f55dd6b4c5effb6fe80f393a55cd9e9d0d5ee3d52c3ac96a1706f9e4566b88dd
+size 10004842440

model-00021-of-00772.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:be0fcadcfb331231b12d8f71260dca722f0999aba10ad4bcaaa03da6b1d4cdbb
+size 10004842440

model-00022-of-00772.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e3b2acf80154609a97b97f7179904a84a1dea91d6c51c6b861ce54dfde5f03b9
+size 10004844152

model-00023-of-00772.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ed66ce25ca4c9a4eac7c1643781ebff98e38737996a5208a05bc50f2c5158b76
+size 10007447584

model-00024-of-00772.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f775b842b9b1d8a4dc812215ae4a069fa3ffee586066f6ff0aa441083489d9fc
+size 10004842232

model-00025-of-00772.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f970819d6b26ce9bd14f246652eb39c73cbcc34a7823edb49c8e32ebf65a7456
+size 10004842440

model-00026-of-00772.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:197e28c697309664e4746a9e079544b740b6b888bed5acd5fd418e2f2f57ec81
+size 10004842440

model-00027-of-00772.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:50ba40a1202e4981c91868d2c883782293f6de087e97065903f29cc44a798f18
+size 10004842440

model-00028-of-00772.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c13ff5c2938fd68dc2be77cf364e9c50594fc63d81b2f68eaf39ebc6d1289748
+size 10004842440

model-00029-of-00772.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7f36902bbfa5495fe00e0b6cdc6f635d1aadd688fce9d5332ccf2edad8a3aec0
+size 10004843768

model-00030-of-00772.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b500fe196b13092815588d27ec8c864d39d3b99f0bea513445014311e080259a
+size 10007448352

model-00031-of-00772.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7e18c20d8a0487cf86e6df70105eb247f3c4c4c7e12f9dd5f5d5e409b2e1db1d
+size 10004841848

model-00032-of-00772.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:188226a1dbdaf6581e81e263dd018851884a381846b4af7f9157af5b3669c80c
+size 10004842440

model-00033-of-00772.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:354992260405316b324acb0933403240a0c8350cd08af05de83cf2a8283456ac
+size 10004842440

model-00034-of-00772.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ac18c6cea682723ea9f01724b0ae012278221aee84ebdc0739e04fc073f85971
+size 10004842440

model-00035-of-00772.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6b770b1c5127fbb7d4d40e47c01f6ff46c7bce6ff735f5f8757490741ecf1b7a
+size 10004842440

model-00036-of-00772.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d85db306d84323f4390978c75b32f2add21c973d5027d5d18cbedb0c0f6bfd6b
+size 10004843384

model-00037-of-00772.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:059b51b172dc72332a7e9ed99549443bd33b692030446bca32e90d6acabc8a4e
+size 10008881016

model-00038-of-00772.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7f11af5e03001b8bf6234f5e9b93a654373e14a812516e7923d463acc3637dd7
+size 10004841496

model-00039-of-00772.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d4cbdaafda955dada32ade9b744d7208567439e587ed11bbdb22388cc22dc25f
+size 10004842440

model-00040-of-00772.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0a6d11579df2164a2bb7fc849152631ee6a87146e39928e167d9e77f5b630d07
+size 10004842440

model-00041-of-00772.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1fbd9b2652c1ddce7cbed1d35bf330829c7e2da90f76aabeeab88d120c12494b
+size 10004842440

model-00042-of-00772.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e63e577eddc17b70ebc678a8a0a8f7b4a57a3558ee2afbb9311461141085a2c9
+size 10004842440

model-00043-of-00772.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ed9df9cef08f7dab6c961c6056e3b8b0451d5500a78706e7537737125db5672d
+size 10004843032