Upload 5 files

config.json

@@ -0,0 +1,32 @@
+{
+  "architectures": [
+    "LlamaForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": 0,
+  "dtype": "bfloat16",
+  "eos_token_id": 0,
+  "head_dim": 64,
+  "hidden_act": "silu",
+  "hidden_size": 2048,
+  "initializer_range": 0.02,
+  "intermediate_size": 8192,
+  "max_position_embeddings": 2048,
+  "mlp_bias": false,
+  "model_type": "llama",
+  "num_attention_heads": 32,
+  "num_hidden_layers": 24,
+  "num_key_value_heads": 32,
+  "pad_token_id": null,
+  "pretraining_tp": 1,
+  "rms_norm_eps": 1e-05,
+  "rope_parameters": {
+    "rope_theta": 10000.0,
+    "rope_type": "default"
+  },
+  "tie_word_embeddings": true,
+  "transformers_version": "5.5.4",
+  "use_cache": true,
+  "vocab_size": 49152
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fe3f5a1b60b7a6085d81cfd74b905c3551f58c34cb32ca898cafc2a785880256
+size 2208640824

spectralq_config.json

@@ -0,0 +1,581 @@
+{
+  "spectralq_version": 1,
+  "dct_layers": [
+    {
+      "name": "model.layers.0.mlp.gate_proj",
+      "out_features": 8192,
+      "in_features": 2048,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 2048
+    },
+    {
+      "name": "model.layers.0.mlp.up_proj",
+      "out_features": 8192,
+      "in_features": 2048,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 2048
+    },
+    {
+      "name": "model.layers.0.mlp.down_proj",
+      "out_features": 2048,
+      "in_features": 8192,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 8192
+    },
+    {
+      "name": "model.layers.1.mlp.gate_proj",
+      "out_features": 8192,
+      "in_features": 2048,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 2048
+    },
+    {
+      "name": "model.layers.1.mlp.up_proj",
+      "out_features": 8192,
+      "in_features": 2048,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 2048
+    },
+    {
+      "name": "model.layers.1.mlp.down_proj",
+      "out_features": 2048,
+      "in_features": 8192,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 8192
+    },
+    {
+      "name": "model.layers.2.mlp.gate_proj",
+      "out_features": 8192,
+      "in_features": 2048,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 2048
+    },
+    {
+      "name": "model.layers.2.mlp.up_proj",
+      "out_features": 8192,
+      "in_features": 2048,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 2048
+    },
+    {
+      "name": "model.layers.2.mlp.down_proj",
+      "out_features": 2048,
+      "in_features": 8192,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 8192
+    },
+    {
+      "name": "model.layers.3.mlp.gate_proj",
+      "out_features": 8192,
+      "in_features": 2048,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 2048
+    },
+    {
+      "name": "model.layers.3.mlp.up_proj",
+      "out_features": 8192,
+      "in_features": 2048,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 2048
+    },
+    {
+      "name": "model.layers.3.mlp.down_proj",
+      "out_features": 2048,
+      "in_features": 8192,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 8192
+    },
+    {
+      "name": "model.layers.4.mlp.gate_proj",
+      "out_features": 8192,
+      "in_features": 2048,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 2048
+    },
+    {
+      "name": "model.layers.4.mlp.up_proj",
+      "out_features": 8192,
+      "in_features": 2048,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 2048
+    },
+    {
+      "name": "model.layers.4.mlp.down_proj",
+      "out_features": 2048,
+      "in_features": 8192,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 8192
+    },
+    {
+      "name": "model.layers.5.mlp.gate_proj",
+      "out_features": 8192,
+      "in_features": 2048,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 2048
+    },
+    {
+      "name": "model.layers.5.mlp.up_proj",
+      "out_features": 8192,
+      "in_features": 2048,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 2048
+    },
+    {
+      "name": "model.layers.5.mlp.down_proj",
+      "out_features": 2048,
+      "in_features": 8192,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 8192
+    },
+    {
+      "name": "model.layers.6.mlp.gate_proj",
+      "out_features": 8192,
+      "in_features": 2048,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 2048
+    },
+    {
+      "name": "model.layers.6.mlp.up_proj",
+      "out_features": 8192,
+      "in_features": 2048,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 2048
+    },
+    {
+      "name": "model.layers.6.mlp.down_proj",
+      "out_features": 2048,
+      "in_features": 8192,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 8192
+    },
+    {
+      "name": "model.layers.7.mlp.gate_proj",
+      "out_features": 8192,
+      "in_features": 2048,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 2048
+    },
+    {
+      "name": "model.layers.7.mlp.up_proj",
+      "out_features": 8192,
+      "in_features": 2048,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 2048
+    },
+    {
+      "name": "model.layers.7.mlp.down_proj",
+      "out_features": 2048,
+      "in_features": 8192,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 8192
+    },
+    {
+      "name": "model.layers.8.mlp.gate_proj",
+      "out_features": 8192,
+      "in_features": 2048,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 2048
+    },
+    {
+      "name": "model.layers.8.mlp.up_proj",
+      "out_features": 8192,
+      "in_features": 2048,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 2048
+    },
+    {
+      "name": "model.layers.8.mlp.down_proj",
+      "out_features": 2048,
+      "in_features": 8192,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 8192
+    },
+    {
+      "name": "model.layers.9.mlp.gate_proj",
+      "out_features": 8192,
+      "in_features": 2048,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 2048
+    },
+    {
+      "name": "model.layers.9.mlp.up_proj",
+      "out_features": 8192,
+      "in_features": 2048,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 2048
+    },
+    {
+      "name": "model.layers.9.mlp.down_proj",
+      "out_features": 2048,
+      "in_features": 8192,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 8192
+    },
+    {
+      "name": "model.layers.10.mlp.gate_proj",
+      "out_features": 8192,
+      "in_features": 2048,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 2048
+    },
+    {
+      "name": "model.layers.10.mlp.up_proj",
+      "out_features": 8192,
+      "in_features": 2048,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 2048
+    },
+    {
+      "name": "model.layers.10.mlp.down_proj",
+      "out_features": 2048,
+      "in_features": 8192,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 8192
+    },
+    {
+      "name": "model.layers.11.mlp.gate_proj",
+      "out_features": 8192,
+      "in_features": 2048,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 2048
+    },
+    {
+      "name": "model.layers.11.mlp.up_proj",
+      "out_features": 8192,
+      "in_features": 2048,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 2048
+    },
+    {
+      "name": "model.layers.11.mlp.down_proj",
+      "out_features": 2048,
+      "in_features": 8192,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 8192
+    },
+    {
+      "name": "model.layers.12.mlp.gate_proj",
+      "out_features": 8192,
+      "in_features": 2048,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 2048
+    },
+    {
+      "name": "model.layers.12.mlp.up_proj",
+      "out_features": 8192,
+      "in_features": 2048,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 2048
+    },
+    {
+      "name": "model.layers.12.mlp.down_proj",
+      "out_features": 2048,
+      "in_features": 8192,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 8192
+    },
+    {
+      "name": "model.layers.13.mlp.gate_proj",
+      "out_features": 8192,
+      "in_features": 2048,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 2048
+    },
+    {
+      "name": "model.layers.13.mlp.up_proj",
+      "out_features": 8192,
+      "in_features": 2048,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 2048
+    },
+    {
+      "name": "model.layers.13.mlp.down_proj",
+      "out_features": 2048,
+      "in_features": 8192,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 8192
+    },
+    {
+      "name": "model.layers.14.mlp.gate_proj",
+      "out_features": 8192,
+      "in_features": 2048,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 2048
+    },
+    {
+      "name": "model.layers.14.mlp.up_proj",
+      "out_features": 8192,
+      "in_features": 2048,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 2048
+    },
+    {
+      "name": "model.layers.14.mlp.down_proj",
+      "out_features": 2048,
+      "in_features": 8192,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 8192
+    },
+    {
+      "name": "model.layers.15.mlp.gate_proj",
+      "out_features": 8192,
+      "in_features": 2048,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 2048
+    },
+    {
+      "name": "model.layers.15.mlp.up_proj",
+      "out_features": 8192,
+      "in_features": 2048,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 2048
+    },
+    {
+      "name": "model.layers.15.mlp.down_proj",
+      "out_features": 2048,
+      "in_features": 8192,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 8192
+    },
+    {
+      "name": "model.layers.16.mlp.gate_proj",
+      "out_features": 8192,
+      "in_features": 2048,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 2048
+    },
+    {
+      "name": "model.layers.16.mlp.up_proj",
+      "out_features": 8192,
+      "in_features": 2048,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 2048
+    },
+    {
+      "name": "model.layers.16.mlp.down_proj",
+      "out_features": 2048,
+      "in_features": 8192,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 8192
+    },
+    {
+      "name": "model.layers.17.mlp.gate_proj",
+      "out_features": 8192,
+      "in_features": 2048,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 2048
+    },
+    {
+      "name": "model.layers.17.mlp.up_proj",
+      "out_features": 8192,
+      "in_features": 2048,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 2048
+    },
+    {
+      "name": "model.layers.17.mlp.down_proj",
+      "out_features": 2048,
+      "in_features": 8192,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 8192
+    },
+    {
+      "name": "model.layers.18.mlp.gate_proj",
+      "out_features": 8192,
+      "in_features": 2048,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 2048
+    },
+    {
+      "name": "model.layers.18.mlp.up_proj",
+      "out_features": 8192,
+      "in_features": 2048,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 2048
+    },
+    {
+      "name": "model.layers.18.mlp.down_proj",
+      "out_features": 2048,
+      "in_features": 8192,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 8192
+    },
+    {
+      "name": "model.layers.19.mlp.gate_proj",
+      "out_features": 8192,
+      "in_features": 2048,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 2048
+    },
+    {
+      "name": "model.layers.19.mlp.up_proj",
+      "out_features": 8192,
+      "in_features": 2048,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 2048
+    },
+    {
+      "name": "model.layers.19.mlp.down_proj",
+      "out_features": 2048,
+      "in_features": 8192,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 8192
+    },
+    {
+      "name": "model.layers.20.mlp.gate_proj",
+      "out_features": 8192,
+      "in_features": 2048,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 2048
+    },
+    {
+      "name": "model.layers.20.mlp.up_proj",
+      "out_features": 8192,
+      "in_features": 2048,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 2048
+    },
+    {
+      "name": "model.layers.20.mlp.down_proj",
+      "out_features": 2048,
+      "in_features": 8192,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 8192
+    },
+    {
+      "name": "model.layers.21.mlp.gate_proj",
+      "out_features": 8192,
+      "in_features": 2048,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 2048
+    },
+    {
+      "name": "model.layers.21.mlp.up_proj",
+      "out_features": 8192,
+      "in_features": 2048,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 2048
+    },
+    {
+      "name": "model.layers.21.mlp.down_proj",
+      "out_features": 2048,
+      "in_features": 8192,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 8192
+    },
+    {
+      "name": "model.layers.22.mlp.gate_proj",
+      "out_features": 8192,
+      "in_features": 2048,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 2048
+    },
+    {
+      "name": "model.layers.22.mlp.up_proj",
+      "out_features": 8192,
+      "in_features": 2048,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 2048
+    },
+    {
+      "name": "model.layers.22.mlp.down_proj",
+      "out_features": 2048,
+      "in_features": 8192,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 8192
+    },
+    {
+      "name": "model.layers.23.mlp.gate_proj",
+      "out_features": 8192,
+      "in_features": 2048,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 2048
+    },
+    {
+      "name": "model.layers.23.mlp.up_proj",
+      "out_features": 8192,
+      "in_features": 2048,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 2048
+    },
+    {
+      "name": "model.layers.23.mlp.down_proj",
+      "out_features": 2048,
+      "in_features": 8192,
+      "block_size": 256,
+      "coeff_bits": 6,
+      "padded_in_f": 8192
+    }
+  ]
+}

tokenizer_config.json

@@ -0,0 +1,33 @@
+{
+  "add_prefix_space": false,
+  "backend": "tokenizers",
+  "bos_token": "<|endoftext|>",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|endoftext|>",
+  "errors": "replace",
+  "extra_special_tokens": [
+    "<|endoftext|>",
+    "<|im_start|>",
+    "<|im_end|>",
+    "<repo_name>",
+    "<reponame>",
+    "<file_sep>",
+    "<filename>",
+    "<gh_stars>",
+    "<issue_start>",
+    "<issue_comment>",
+    "<issue_closed>",
+    "<jupyter_start>",
+    "<jupyter_text>",
+    "<jupyter_code>",
+    "<jupyter_output>",
+    "<jupyter_script>",
+    "<empty_output>"
+  ],
+  "is_local": false,
+  "model_max_length": 2048,
+  "pad_token": null,
+  "tokenizer_class": "GPT2Tokenizer",
+  "unk_token": "<|endoftext|>",
+  "vocab_size": 49152
+}