Upload folder using huggingface_hub

layouts/AesCode-8B-IQ2_M.layout.json

@@ -0,0 +1,90 @@
+{
+  "model": "/media/m2teamgroup/models/AesCode-8B-GGUF/AesCode-8B-bf16.gguf",
+  "imatrix": "/media/m2teamgroup/models/AesCode-8B-GGUF/AesCode-8B-imatrix.gguf",
+  "ftype": "IQ2_M",
+  "base_type": "iq2_s",
+  "family": "iq",
+  "recipe": "ident",
+  "variant": "shr0.70",
+  "auto_crush": false,
+  "rung_mode": "share",
+  "tier": "M",
+  "rung_share": 0.7,
+  "crush_pass_bytes": null,
+  "embd_shares": {
+    "output.weight": {
+      "share": 0.1274,
+      "type": "q5_k"
+    },
+    "token_embd.weight": {
+      "share": 0.0796,
+      "type": "q3_k"
+    }
+  },
+  "arch": "dense",
+  "tied": false,
+  "n_layer": 36,
+  "body_rule": true,
+  "pins": {},
+  "shape_pins": {},
+  "fixed_pins": {},
+  "embeddings": {
+    "output.weight": "q5_k",
+    "token_embd.weight": "q3_k"
+  },
+  "bytes": {
+    "heuristic_body+embd": 3357134848,
+    "map_body+embd": 3132706816,
+    "ratio": 0.9331,
+    "note": "quantizable, unpinned tensors only; pinned/unquantized tensors are identical in both"
+  },
+  "body_base_share": 0.7,
+  "body_bpw": 2.807,
+  "body_bpw_heuristic": 3.066,
+  "total_params": 8190735360,
+  "file_bytes_pred": 3133939712,
+  "file_bpw_pred": 3.061,
+  "histogram": {
+    "attn_k": {
+      "iq4_xs": 27,
+      "iq2_s": 6,
+      "iq3_s": 2,
+      "iq3_xxs": 1
+    },
+    "attn_output": {
+      "iq3_xxs": 17,
+      "iq2_s": 10,
+      "iq4_xs": 8,
+      "iq3_s": 1
+    },
+    "attn_q": {
+      "iq2_s": 36
+    },
+    "attn_v": {
+      "q4_k": 27,
+      "iq4_xs": 9
+    },
+    "ffn_down": {
+      "iq2_s": 27,
+      "iq3_s": 9
+    },
+    "ffn_gate": {
+      "iq2_s": 36
+    },
+    "ffn_up": {
+      "iq2_s": 27,
+      "iq3_s": 8,
+      "iq4_xs": 1
+    }
+  },
+  "prior": "layout/prior.json",
+  "pin_rule": "floor",
+  "embd_cap": "heur",
+  "generator": "auto_quant_v2.layout",
+  "generator_key": "3bf8b43e20a68f0b",
+  "quant_type": "IQ2_M",
+  "quantize_ftype": "IQ2_M",
+  "llama_cpp_version": "b11490",
+  "unknown_kinds": {},
+  "commit": "5d550ef5561a30307f1d436db2c6835b75395e77"
+}

layouts/AesCode-8B-IQ2_M.tensor-types.txt

@@ -0,0 +1,254 @@
+^output\.weight$=q5_k
+^token_embd\.weight$=q3_k
+^blk\.0\.attn_k\.weight$=iq4_xs
+^blk\.0\.attn_output\.weight$=iq2_s
+^blk\.0\.attn_q\.weight$=iq2_s
+^blk\.0\.attn_v\.weight$=iq4_xs
+^blk\.0\.ffn_down\.weight$=iq2_s
+^blk\.0\.ffn_gate\.weight$=iq2_s
+^blk\.0\.ffn_up\.weight$=iq2_s
+^blk\.1\.attn_k\.weight$=iq4_xs
+^blk\.1\.attn_output\.weight$=iq4_xs
+^blk\.1\.attn_q\.weight$=iq2_s
+^blk\.1\.attn_v\.weight$=q4_k
+^blk\.1\.ffn_down\.weight$=iq3_s
+^blk\.1\.ffn_gate\.weight$=iq2_s
+^blk\.1\.ffn_up\.weight$=iq3_s
+^blk\.2\.attn_k\.weight$=iq4_xs
+^blk\.2\.attn_output\.weight$=iq4_xs
+^blk\.2\.attn_q\.weight$=iq2_s
+^blk\.2\.attn_v\.weight$=q4_k
+^blk\.2\.ffn_down\.weight$=iq3_s
+^blk\.2\.ffn_gate\.weight$=iq2_s
+^blk\.2\.ffn_up\.weight$=iq3_s
+^blk\.3\.attn_k\.weight$=iq4_xs
+^blk\.3\.attn_output\.weight$=iq4_xs
+^blk\.3\.attn_q\.weight$=iq2_s
+^blk\.3\.attn_v\.weight$=q4_k
+^blk\.3\.ffn_down\.weight$=iq3_s
+^blk\.3\.ffn_gate\.weight$=iq2_s
+^blk\.3\.ffn_up\.weight$=iq3_s
+^blk\.4\.attn_k\.weight$=iq4_xs
+^blk\.4\.attn_output\.weight$=iq4_xs
+^blk\.4\.attn_q\.weight$=iq2_s
+^blk\.4\.attn_v\.weight$=q4_k
+^blk\.4\.ffn_down\.weight$=iq3_s
+^blk\.4\.ffn_gate\.weight$=iq2_s
+^blk\.4\.ffn_up\.weight$=iq3_s
+^blk\.5\.attn_k\.weight$=iq4_xs
+^blk\.5\.attn_output\.weight$=iq4_xs
+^blk\.5\.attn_q\.weight$=iq2_s
+^blk\.5\.attn_v\.weight$=q4_k
+^blk\.5\.ffn_down\.weight$=iq3_s
+^blk\.5\.ffn_gate\.weight$=iq2_s
+^blk\.5\.ffn_up\.weight$=iq3_s
+^blk\.6\.attn_k\.weight$=iq4_xs
+^blk\.6\.attn_output\.weight$=iq4_xs
+^blk\.6\.attn_q\.weight$=iq2_s
+^blk\.6\.attn_v\.weight$=q4_k
+^blk\.6\.ffn_down\.weight$=iq3_s
+^blk\.6\.ffn_gate\.weight$=iq2_s
+^blk\.6\.ffn_up\.weight$=iq3_s
+^blk\.7\.attn_k\.weight$=iq4_xs
+^blk\.7\.attn_output\.weight$=iq4_xs
+^blk\.7\.attn_q\.weight$=iq2_s
+^blk\.7\.attn_v\.weight$=q4_k
+^blk\.7\.ffn_down\.weight$=iq3_s
+^blk\.7\.ffn_gate\.weight$=iq2_s
+^blk\.7\.ffn_up\.weight$=iq3_s
+^blk\.8\.attn_k\.weight$=iq4_xs
+^blk\.8\.attn_output\.weight$=iq4_xs
+^blk\.8\.attn_q\.weight$=iq2_s
+^blk\.8\.attn_v\.weight$=q4_k
+^blk\.8\.ffn_down\.weight$=iq3_s
+^blk\.8\.ffn_gate\.weight$=iq2_s
+^blk\.8\.ffn_up\.weight$=iq3_s
+^blk\.9\.attn_k\.weight$=iq4_xs
+^blk\.9\.attn_output\.weight$=iq3_s
+^blk\.9\.attn_q\.weight$=iq2_s
+^blk\.9\.attn_v\.weight$=q4_k
+^blk\.9\.ffn_down\.weight$=iq2_s
+^blk\.9\.ffn_gate\.weight$=iq2_s
+^blk\.9\.ffn_up\.weight$=iq2_s
+^blk\.10\.attn_k\.weight$=iq4_xs
+^blk\.10\.attn_output\.weight$=iq3_xxs
+^blk\.10\.attn_q\.weight$=iq2_s
+^blk\.10\.attn_v\.weight$=q4_k
+^blk\.10\.ffn_down\.weight$=iq2_s
+^blk\.10\.ffn_gate\.weight$=iq2_s
+^blk\.10\.ffn_up\.weight$=iq2_s
+^blk\.11\.attn_k\.weight$=iq4_xs
+^blk\.11\.attn_output\.weight$=iq3_xxs
+^blk\.11\.attn_q\.weight$=iq2_s
+^blk\.11\.attn_v\.weight$=q4_k
+^blk\.11\.ffn_down\.weight$=iq2_s
+^blk\.11\.ffn_gate\.weight$=iq2_s
+^blk\.11\.ffn_up\.weight$=iq2_s
+^blk\.12\.attn_k\.weight$=iq4_xs
+^blk\.12\.attn_output\.weight$=iq3_xxs
+^blk\.12\.attn_q\.weight$=iq2_s
+^blk\.12\.attn_v\.weight$=q4_k
+^blk\.12\.ffn_down\.weight$=iq2_s
+^blk\.12\.ffn_gate\.weight$=iq2_s
+^blk\.12\.ffn_up\.weight$=iq2_s
+^blk\.13\.attn_k\.weight$=iq4_xs
+^blk\.13\.attn_output\.weight$=iq3_xxs
+^blk\.13\.attn_q\.weight$=iq2_s
+^blk\.13\.attn_v\.weight$=q4_k
+^blk\.13\.ffn_down\.weight$=iq2_s
+^blk\.13\.ffn_gate\.weight$=iq2_s
+^blk\.13\.ffn_up\.weight$=iq2_s
+^blk\.14\.attn_k\.weight$=iq4_xs
+^blk\.14\.attn_output\.weight$=iq3_xxs
+^blk\.14\.attn_q\.weight$=iq2_s
+^blk\.14\.attn_v\.weight$=q4_k
+^blk\.14\.ffn_down\.weight$=iq2_s
+^blk\.14\.ffn_gate\.weight$=iq2_s
+^blk\.14\.ffn_up\.weight$=iq2_s
+^blk\.15\.attn_k\.weight$=iq4_xs
+^blk\.15\.attn_output\.weight$=iq3_xxs
+^blk\.15\.attn_q\.weight$=iq2_s
+^blk\.15\.attn_v\.weight$=q4_k
+^blk\.15\.ffn_down\.weight$=iq2_s
+^blk\.15\.ffn_gate\.weight$=iq2_s
+^blk\.15\.ffn_up\.weight$=iq2_s
+^blk\.16\.attn_k\.weight$=iq4_xs
+^blk\.16\.attn_output\.weight$=iq3_xxs
+^blk\.16\.attn_q\.weight$=iq2_s
+^blk\.16\.attn_v\.weight$=q4_k
+^blk\.16\.ffn_down\.weight$=iq2_s
+^blk\.16\.ffn_gate\.weight$=iq2_s
+^blk\.16\.ffn_up\.weight$=iq2_s
+^blk\.17\.attn_k\.weight$=iq4_xs
+^blk\.17\.attn_output\.weight$=iq3_xxs
+^blk\.17\.attn_q\.weight$=iq2_s
+^blk\.17\.attn_v\.weight$=q4_k
+^blk\.17\.ffn_down\.weight$=iq2_s
+^blk\.17\.ffn_gate\.weight$=iq2_s
+^blk\.17\.ffn_up\.weight$=iq2_s
+^blk\.18\.attn_k\.weight$=iq4_xs
+^blk\.18\.attn_output\.weight$=iq3_xxs
+^blk\.18\.attn_q\.weight$=iq2_s
+^blk\.18\.attn_v\.weight$=q4_k
+^blk\.18\.ffn_down\.weight$=iq2_s
+^blk\.18\.ffn_gate\.weight$=iq2_s
+^blk\.18\.ffn_up\.weight$=iq2_s
+^blk\.19\.attn_k\.weight$=iq4_xs
+^blk\.19\.attn_output\.weight$=iq3_xxs
+^blk\.19\.attn_q\.weight$=iq2_s
+^blk\.19\.attn_v\.weight$=q4_k
+^blk\.19\.ffn_down\.weight$=iq2_s
+^blk\.19\.ffn_gate\.weight$=iq2_s
+^blk\.19\.ffn_up\.weight$=iq2_s
+^blk\.20\.attn_k\.weight$=iq4_xs
+^blk\.20\.attn_output\.weight$=iq3_xxs
+^blk\.20\.attn_q\.weight$=iq2_s
+^blk\.20\.attn_v\.weight$=q4_k
+^blk\.20\.ffn_down\.weight$=iq2_s
+^blk\.20\.ffn_gate\.weight$=iq2_s
+^blk\.20\.ffn_up\.weight$=iq2_s
+^blk\.21\.attn_k\.weight$=iq4_xs
+^blk\.21\.attn_output\.weight$=iq3_xxs
+^blk\.21\.attn_q\.weight$=iq2_s
+^blk\.21\.attn_v\.weight$=q4_k
+^blk\.21\.ffn_down\.weight$=iq2_s
+^blk\.21\.ffn_gate\.weight$=iq2_s
+^blk\.21\.ffn_up\.weight$=iq2_s
+^blk\.22\.attn_k\.weight$=iq4_xs
+^blk\.22\.attn_output\.weight$=iq3_xxs
+^blk\.22\.attn_q\.weight$=iq2_s
+^blk\.22\.attn_v\.weight$=q4_k
+^blk\.22\.ffn_down\.weight$=iq2_s
+^blk\.22\.ffn_gate\.weight$=iq2_s
+^blk\.22\.ffn_up\.weight$=iq2_s
+^blk\.23\.attn_k\.weight$=iq4_xs
+^blk\.23\.attn_output\.weight$=iq3_xxs
+^blk\.23\.attn_q\.weight$=iq2_s
+^blk\.23\.attn_v\.weight$=q4_k
+^blk\.23\.ffn_down\.weight$=iq2_s
+^blk\.23\.ffn_gate\.weight$=iq2_s
+^blk\.23\.ffn_up\.weight$=iq2_s
+^blk\.24\.attn_k\.weight$=iq4_xs
+^blk\.24\.attn_output\.weight$=iq3_xxs
+^blk\.24\.attn_q\.weight$=iq2_s
+^blk\.24\.attn_v\.weight$=q4_k
+^blk\.24\.ffn_down\.weight$=iq2_s
+^blk\.24\.ffn_gate\.weight$=iq2_s
+^blk\.24\.ffn_up\.weight$=iq2_s
+^blk\.25\.attn_k\.weight$=iq4_xs
+^blk\.25\.attn_output\.weight$=iq3_xxs
+^blk\.25\.attn_q\.weight$=iq2_s
+^blk\.25\.attn_v\.weight$=q4_k
+^blk\.25\.ffn_down\.weight$=iq2_s
+^blk\.25\.ffn_gate\.weight$=iq2_s
+^blk\.25\.ffn_up\.weight$=iq2_s
+^blk\.26\.attn_k\.weight$=iq4_xs
+^blk\.26\.attn_output\.weight$=iq3_xxs
+^blk\.26\.attn_q\.weight$=iq2_s
+^blk\.26\.attn_v\.weight$=q4_k
+^blk\.26\.ffn_down\.weight$=iq2_s
+^blk\.26\.ffn_gate\.weight$=iq2_s
+^blk\.26\.ffn_up\.weight$=iq2_s
+^blk\.27\.attn_k\.weight$=iq3_s
+^blk\.27\.attn_output\.weight$=iq2_s
+^blk\.27\.attn_q\.weight$=iq2_s
+^blk\.27\.attn_v\.weight$=iq4_xs
+^blk\.27\.ffn_down\.weight$=iq2_s
+^blk\.27\.ffn_gate\.weight$=iq2_s
+^blk\.27\.ffn_up\.weight$=iq2_s
+^blk\.28\.attn_k\.weight$=iq3_s
+^blk\.28\.attn_output\.weight$=iq2_s
+^blk\.28\.attn_q\.weight$=iq2_s
+^blk\.28\.attn_v\.weight$=iq4_xs
+^blk\.28\.ffn_down\.weight$=iq2_s
+^blk\.28\.ffn_gate\.weight$=iq2_s
+^blk\.28\.ffn_up\.weight$=iq2_s
+^blk\.29\.attn_k\.weight$=iq3_xxs
+^blk\.29\.attn_output\.weight$=iq2_s
+^blk\.29\.attn_q\.weight$=iq2_s
+^blk\.29\.attn_v\.weight$=iq4_xs
+^blk\.29\.ffn_down\.weight$=iq2_s
+^blk\.29\.ffn_gate\.weight$=iq2_s
+^blk\.29\.ffn_up\.weight$=iq2_s
+^blk\.30\.attn_k\.weight$=iq2_s
+^blk\.30\.attn_output\.weight$=iq2_s
+^blk\.30\.attn_q\.weight$=iq2_s
+^blk\.30\.attn_v\.weight$=iq4_xs
+^blk\.30\.ffn_down\.weight$=iq2_s
+^blk\.30\.ffn_gate\.weight$=iq2_s
+^blk\.30\.ffn_up\.weight$=iq2_s
+^blk\.31\.attn_k\.weight$=iq2_s
+^blk\.31\.attn_output\.weight$=iq2_s
+^blk\.31\.attn_q\.weight$=iq2_s
+^blk\.31\.attn_v\.weight$=iq4_xs
+^blk\.31\.ffn_down\.weight$=iq2_s
+^blk\.31\.ffn_gate\.weight$=iq2_s
+^blk\.31\.ffn_up\.weight$=iq2_s
+^blk\.32\.attn_k\.weight$=iq2_s
+^blk\.32\.attn_output\.weight$=iq2_s
+^blk\.32\.attn_q\.weight$=iq2_s
+^blk\.32\.attn_v\.weight$=iq4_xs
+^blk\.32\.ffn_down\.weight$=iq2_s
+^blk\.32\.ffn_gate\.weight$=iq2_s
+^blk\.32\.ffn_up\.weight$=iq2_s
+^blk\.33\.attn_k\.weight$=iq2_s
+^blk\.33\.attn_output\.weight$=iq2_s
+^blk\.33\.attn_q\.weight$=iq2_s
+^blk\.33\.attn_v\.weight$=iq4_xs
+^blk\.33\.ffn_down\.weight$=iq2_s
+^blk\.33\.ffn_gate\.weight$=iq2_s
+^blk\.33\.ffn_up\.weight$=iq2_s
+^blk\.34\.attn_k\.weight$=iq2_s
+^blk\.34\.attn_output\.weight$=iq2_s
+^blk\.34\.attn_q\.weight$=iq2_s
+^blk\.34\.attn_v\.weight$=iq4_xs
+^blk\.34\.ffn_down\.weight$=iq2_s
+^blk\.34\.ffn_gate\.weight$=iq2_s
+^blk\.34\.ffn_up\.weight$=iq2_s
+^blk\.35\.attn_k\.weight$=iq2_s
+^blk\.35\.attn_output\.weight$=iq2_s
+^blk\.35\.attn_q\.weight$=iq2_s
+^blk\.35\.attn_v\.weight$=q4_k
+^blk\.35\.ffn_down\.weight$=iq3_s
+^blk\.35\.ffn_gate\.weight$=iq2_s
+^blk\.35\.ffn_up\.weight$=iq4_xs

layouts/AesCode-8B-IQ3_M.layout.json

@@ -0,0 +1,88 @@
+{
+  "model": "/media/m2teamgroup/models/AesCode-8B-GGUF/AesCode-8B-bf16.gguf",
+  "imatrix": "/media/m2teamgroup/models/AesCode-8B-GGUF/AesCode-8B-imatrix.gguf",
+  "ftype": "IQ3_M",
+  "base_type": "iq3_s",
+  "family": "iq",
+  "recipe": "ident",
+  "variant": "shr0.50",
+  "auto_crush": false,
+  "rung_mode": "share",
+  "tier": "L",
+  "rung_share": 0.5,
+  "crush_pass_bytes": null,
+  "embd_shares": {
+    "output.weight": {
+      "share": 0.1268,
+      "type": "q5_k"
+    },
+    "token_embd.weight": {
+      "share": 0.0664,
+      "type": "q4_k"
+    }
+  },
+  "arch": "dense",
+  "tied": false,
+  "n_layer": 36,
+  "body_rule": true,
+  "pins": {},
+  "shape_pins": {},
+  "fixed_pins": {},
+  "embeddings": {
+    "output.weight": "q5_k",
+    "token_embd.weight": "q4_k"
+  },
+  "bytes": {
+    "heuristic_body+embd": 4026269696,
+    "map_body+embd": 4468572160,
+    "ratio": 1.1099,
+    "note": "quantizable, unpinned tensors only; pinned/unquantized tensors are identical in both"
+  },
+  "body_base_share": 0.5,
+  "body_bpw": 4.251,
+  "body_bpw_heuristic": 3.741,
+  "total_params": 8190735360,
+  "file_bytes_pred": 4469805056,
+  "file_bpw_pred": 4.366,
+  "histogram": {
+    "attn_k": {
+      "q5_k": 35,
+      "q6_k": 1
+    },
+    "attn_output": {
+      "q5_k": 28,
+      "iq3_s": 8
+    },
+    "attn_q": {
+      "iq3_s": 35,
+      "q4_k": 1
+    },
+    "attn_v": {
+      "q6_k": 21,
+      "q8_0": 8,
+      "q5_k": 7
+    },
+    "ffn_down": {
+      "q5_k": 18,
+      "iq3_s": 18
+    },
+    "ffn_gate": {
+      "iq3_s": 35,
+      "q5_k": 1
+    },
+    "ffn_up": {
+      "q5_k": 18,
+      "iq3_s": 18
+    }
+  },
+  "prior": "layout/prior.json",
+  "pin_rule": "floor",
+  "embd_cap": "heur",
+  "generator": "auto_quant_v2.layout",
+  "generator_key": "3bf8b43e20a68f0b",
+  "quant_type": "IQ3_M",
+  "quantize_ftype": "IQ3_M",
+  "llama_cpp_version": "b11490",
+  "unknown_kinds": {},
+  "commit": "5d550ef5561a30307f1d436db2c6835b75395e77"
+}

layouts/AesCode-8B-IQ3_M.tensor-types.txt

@@ -0,0 +1,254 @@
+^output\.weight$=q5_k
+^token_embd\.weight$=q4_k
+^blk\.0\.attn_k\.weight$=q6_k
+^blk\.0\.attn_output\.weight$=q5_k
+^blk\.0\.attn_q\.weight$=iq3_s
+^blk\.0\.attn_v\.weight$=q6_k
+^blk\.0\.ffn_down\.weight$=q5_k
+^blk\.0\.ffn_gate\.weight$=iq3_s
+^blk\.0\.ffn_up\.weight$=q5_k
+^blk\.1\.attn_k\.weight$=q5_k
+^blk\.1\.attn_output\.weight$=q5_k
+^blk\.1\.attn_q\.weight$=iq3_s
+^blk\.1\.attn_v\.weight$=q8_0
+^blk\.1\.ffn_down\.weight$=q5_k
+^blk\.1\.ffn_gate\.weight$=iq3_s
+^blk\.1\.ffn_up\.weight$=q5_k
+^blk\.2\.attn_k\.weight$=q5_k
+^blk\.2\.attn_output\.weight$=q5_k
+^blk\.2\.attn_q\.weight$=iq3_s
+^blk\.2\.attn_v\.weight$=q8_0
+^blk\.2\.ffn_down\.weight$=q5_k
+^blk\.2\.ffn_gate\.weight$=iq3_s
+^blk\.2\.ffn_up\.weight$=q5_k
+^blk\.3\.attn_k\.weight$=q5_k
+^blk\.3\.attn_output\.weight$=q5_k
+^blk\.3\.attn_q\.weight$=iq3_s
+^blk\.3\.attn_v\.weight$=q8_0
+^blk\.3\.ffn_down\.weight$=q5_k
+^blk\.3\.ffn_gate\.weight$=iq3_s
+^blk\.3\.ffn_up\.weight$=q5_k
+^blk\.4\.attn_k\.weight$=q5_k
+^blk\.4\.attn_output\.weight$=q5_k
+^blk\.4\.attn_q\.weight$=iq3_s
+^blk\.4\.attn_v\.weight$=q8_0
+^blk\.4\.ffn_down\.weight$=q5_k
+^blk\.4\.ffn_gate\.weight$=iq3_s
+^blk\.4\.ffn_up\.weight$=q5_k
+^blk\.5\.attn_k\.weight$=q5_k
+^blk\.5\.attn_output\.weight$=q5_k
+^blk\.5\.attn_q\.weight$=iq3_s
+^blk\.5\.attn_v\.weight$=q8_0
+^blk\.5\.ffn_down\.weight$=q5_k
+^blk\.5\.ffn_gate\.weight$=iq3_s
+^blk\.5\.ffn_up\.weight$=q5_k
+^blk\.6\.attn_k\.weight$=q5_k
+^blk\.6\.attn_output\.weight$=q5_k
+^blk\.6\.attn_q\.weight$=iq3_s
+^blk\.6\.attn_v\.weight$=q8_0
+^blk\.6\.ffn_down\.weight$=q5_k
+^blk\.6\.ffn_gate\.weight$=iq3_s
+^blk\.6\.ffn_up\.weight$=q5_k
+^blk\.7\.attn_k\.weight$=q5_k
+^blk\.7\.attn_output\.weight$=q5_k
+^blk\.7\.attn_q\.weight$=iq3_s
+^blk\.7\.attn_v\.weight$=q8_0
+^blk\.7\.ffn_down\.weight$=q5_k
+^blk\.7\.ffn_gate\.weight$=iq3_s
+^blk\.7\.ffn_up\.weight$=q5_k
+^blk\.8\.attn_k\.weight$=q5_k
+^blk\.8\.attn_output\.weight$=q5_k
+^blk\.8\.attn_q\.weight$=iq3_s
+^blk\.8\.attn_v\.weight$=q8_0
+^blk\.8\.ffn_down\.weight$=q5_k
+^blk\.8\.ffn_gate\.weight$=iq3_s
+^blk\.8\.ffn_up\.weight$=q5_k
+^blk\.9\.attn_k\.weight$=q5_k
+^blk\.9\.attn_output\.weight$=q5_k
+^blk\.9\.attn_q\.weight$=q4_k
+^blk\.9\.attn_v\.weight$=q6_k
+^blk\.9\.ffn_down\.weight$=iq3_s
+^blk\.9\.ffn_gate\.weight$=iq3_s
+^blk\.9\.ffn_up\.weight$=iq3_s
+^blk\.10\.attn_k\.weight$=q5_k
+^blk\.10\.attn_output\.weight$=q5_k
+^blk\.10\.attn_q\.weight$=iq3_s
+^blk\.10\.attn_v\.weight$=q6_k
+^blk\.10\.ffn_down\.weight$=iq3_s
+^blk\.10\.ffn_gate\.weight$=iq3_s
+^blk\.10\.ffn_up\.weight$=iq3_s
+^blk\.11\.attn_k\.weight$=q5_k
+^blk\.11\.attn_output\.weight$=q5_k
+^blk\.11\.attn_q\.weight$=iq3_s
+^blk\.11\.attn_v\.weight$=q6_k
+^blk\.11\.ffn_down\.weight$=iq3_s
+^blk\.11\.ffn_gate\.weight$=iq3_s
+^blk\.11\.ffn_up\.weight$=iq3_s
+^blk\.12\.attn_k\.weight$=q5_k
+^blk\.12\.attn_output\.weight$=q5_k
+^blk\.12\.attn_q\.weight$=iq3_s
+^blk\.12\.attn_v\.weight$=q6_k
+^blk\.12\.ffn_down\.weight$=iq3_s
+^blk\.12\.ffn_gate\.weight$=iq3_s
+^blk\.12\.ffn_up\.weight$=iq3_s
+^blk\.13\.attn_k\.weight$=q5_k
+^blk\.13\.attn_output\.weight$=q5_k
+^blk\.13\.attn_q\.weight$=iq3_s
+^blk\.13\.attn_v\.weight$=q6_k
+^blk\.13\.ffn_down\.weight$=iq3_s
+^blk\.13\.ffn_gate\.weight$=iq3_s
+^blk\.13\.ffn_up\.weight$=iq3_s
+^blk\.14\.attn_k\.weight$=q5_k
+^blk\.14\.attn_output\.weight$=q5_k
+^blk\.14\.attn_q\.weight$=iq3_s
+^blk\.14\.attn_v\.weight$=q6_k
+^blk\.14\.ffn_down\.weight$=iq3_s
+^blk\.14\.ffn_gate\.weight$=iq3_s
+^blk\.14\.ffn_up\.weight$=iq3_s
+^blk\.15\.attn_k\.weight$=q5_k
+^blk\.15\.attn_output\.weight$=q5_k
+^blk\.15\.attn_q\.weight$=iq3_s
+^blk\.15\.attn_v\.weight$=q6_k
+^blk\.15\.ffn_down\.weight$=iq3_s
+^blk\.15\.ffn_gate\.weight$=iq3_s
+^blk\.15\.ffn_up\.weight$=iq3_s
+^blk\.16\.attn_k\.weight$=q5_k
+^blk\.16\.attn_output\.weight$=q5_k
+^blk\.16\.attn_q\.weight$=iq3_s
+^blk\.16\.attn_v\.weight$=q6_k
+^blk\.16\.ffn_down\.weight$=iq3_s
+^blk\.16\.ffn_gate\.weight$=iq3_s
+^blk\.16\.ffn_up\.weight$=iq3_s
+^blk\.17\.attn_k\.weight$=q5_k
+^blk\.17\.attn_output\.weight$=q5_k
+^blk\.17\.attn_q\.weight$=iq3_s
+^blk\.17\.attn_v\.weight$=q6_k
+^blk\.17\.ffn_down\.weight$=iq3_s
+^blk\.17\.ffn_gate\.weight$=iq3_s
+^blk\.17\.ffn_up\.weight$=iq3_s
+^blk\.18\.attn_k\.weight$=q5_k
+^blk\.18\.attn_output\.weight$=q5_k
+^blk\.18\.attn_q\.weight$=iq3_s
+^blk\.18\.attn_v\.weight$=q6_k
+^blk\.18\.ffn_down\.weight$=iq3_s
+^blk\.18\.ffn_gate\.weight$=iq3_s
+^blk\.18\.ffn_up\.weight$=iq3_s
+^blk\.19\.attn_k\.weight$=q5_k
+^blk\.19\.attn_output\.weight$=q5_k
+^blk\.19\.attn_q\.weight$=iq3_s
+^blk\.19\.attn_v\.weight$=q6_k
+^blk\.19\.ffn_down\.weight$=iq3_s
+^blk\.19\.ffn_gate\.weight$=iq3_s
+^blk\.19\.ffn_up\.weight$=iq3_s
+^blk\.20\.attn_k\.weight$=q5_k
+^blk\.20\.attn_output\.weight$=q5_k
+^blk\.20\.attn_q\.weight$=iq3_s
+^blk\.20\.attn_v\.weight$=q6_k
+^blk\.20\.ffn_down\.weight$=iq3_s
+^blk\.20\.ffn_gate\.weight$=iq3_s
+^blk\.20\.ffn_up\.weight$=iq3_s
+^blk\.21\.attn_k\.weight$=q5_k
+^blk\.21\.attn_output\.weight$=q5_k
+^blk\.21\.attn_q\.weight$=iq3_s
+^blk\.21\.attn_v\.weight$=q6_k
+^blk\.21\.ffn_down\.weight$=iq3_s
+^blk\.21\.ffn_gate\.weight$=iq3_s
+^blk\.21\.ffn_up\.weight$=iq3_s
+^blk\.22\.attn_k\.weight$=q5_k
+^blk\.22\.attn_output\.weight$=q5_k
+^blk\.22\.attn_q\.weight$=iq3_s
+^blk\.22\.attn_v\.weight$=q6_k
+^blk\.22\.ffn_down\.weight$=iq3_s
+^blk\.22\.ffn_gate\.weight$=iq3_s
+^blk\.22\.ffn_up\.weight$=iq3_s
+^blk\.23\.attn_k\.weight$=q5_k
+^blk\.23\.attn_output\.weight$=q5_k
+^blk\.23\.attn_q\.weight$=iq3_s
+^blk\.23\.attn_v\.weight$=q6_k
+^blk\.23\.ffn_down\.weight$=iq3_s
+^blk\.23\.ffn_gate\.weight$=iq3_s
+^blk\.23\.ffn_up\.weight$=iq3_s
+^blk\.24\.attn_k\.weight$=q5_k
+^blk\.24\.attn_output\.weight$=q5_k
+^blk\.24\.attn_q\.weight$=iq3_s
+^blk\.24\.attn_v\.weight$=q6_k
+^blk\.24\.ffn_down\.weight$=iq3_s
+^blk\.24\.ffn_gate\.weight$=iq3_s
+^blk\.24\.ffn_up\.weight$=iq3_s
+^blk\.25\.attn_k\.weight$=q5_k
+^blk\.25\.attn_output\.weight$=q5_k
+^blk\.25\.attn_q\.weight$=iq3_s
+^blk\.25\.attn_v\.weight$=q6_k
+^blk\.25\.ffn_down\.weight$=iq3_s
+^blk\.25\.ffn_gate\.weight$=iq3_s
+^blk\.25\.ffn_up\.weight$=iq3_s
+^blk\.26\.attn_k\.weight$=q5_k
+^blk\.26\.attn_output\.weight$=q5_k
+^blk\.26\.attn_q\.weight$=iq3_s
+^blk\.26\.attn_v\.weight$=q6_k
+^blk\.26\.ffn_down\.weight$=iq3_s
+^blk\.26\.ffn_gate\.weight$=iq3_s
+^blk\.26\.ffn_up\.weight$=iq3_s
+^blk\.27\.attn_k\.weight$=q5_k
+^blk\.27\.attn_output\.weight$=iq3_s
+^blk\.27\.attn_q\.weight$=iq3_s
+^blk\.27\.attn_v\.weight$=q6_k
+^blk\.27\.ffn_down\.weight$=q5_k
+^blk\.27\.ffn_gate\.weight$=iq3_s
+^blk\.27\.ffn_up\.weight$=q5_k
+^blk\.28\.attn_k\.weight$=q5_k
+^blk\.28\.attn_output\.weight$=iq3_s
+^blk\.28\.attn_q\.weight$=iq3_s
+^blk\.28\.attn_v\.weight$=q5_k
+^blk\.28\.ffn_down\.weight$=q5_k
+^blk\.28\.ffn_gate\.weight$=iq3_s
+^blk\.28\.ffn_up\.weight$=q5_k
+^blk\.29\.attn_k\.weight$=q5_k
+^blk\.29\.attn_output\.weight$=iq3_s
+^blk\.29\.attn_q\.weight$=iq3_s
+^blk\.29\.attn_v\.weight$=q5_k
+^blk\.29\.ffn_down\.weight$=q5_k
+^blk\.29\.ffn_gate\.weight$=iq3_s
+^blk\.29\.ffn_up\.weight$=q5_k
+^blk\.30\.attn_k\.weight$=q5_k
+^blk\.30\.attn_output\.weight$=iq3_s
+^blk\.30\.attn_q\.weight$=iq3_s
+^blk\.30\.attn_v\.weight$=q5_k
+^blk\.30\.ffn_down\.weight$=q5_k
+^blk\.30\.ffn_gate\.weight$=iq3_s
+^blk\.30\.ffn_up\.weight$=q5_k
+^blk\.31\.attn_k\.weight$=q5_k
+^blk\.31\.attn_output\.weight$=iq3_s
+^blk\.31\.attn_q\.weight$=iq3_s
+^blk\.31\.attn_v\.weight$=q5_k
+^blk\.31\.ffn_down\.weight$=q5_k
+^blk\.31\.ffn_gate\.weight$=iq3_s
+^blk\.31\.ffn_up\.weight$=q5_k
+^blk\.32\.attn_k\.weight$=q5_k
+^blk\.32\.attn_output\.weight$=iq3_s
+^blk\.32\.attn_q\.weight$=iq3_s
+^blk\.32\.attn_v\.weight$=q5_k
+^blk\.32\.ffn_down\.weight$=q5_k
+^blk\.32\.ffn_gate\.weight$=iq3_s
+^blk\.32\.ffn_up\.weight$=q5_k
+^blk\.33\.attn_k\.weight$=q5_k
+^blk\.33\.attn_output\.weight$=iq3_s
+^blk\.33\.attn_q\.weight$=iq3_s
+^blk\.33\.attn_v\.weight$=q5_k
+^blk\.33\.ffn_down\.weight$=q5_k
+^blk\.33\.ffn_gate\.weight$=iq3_s
+^blk\.33\.ffn_up\.weight$=q5_k
+^blk\.34\.attn_k\.weight$=q5_k
+^blk\.34\.attn_output\.weight$=iq3_s
+^blk\.34\.attn_q\.weight$=iq3_s
+^blk\.34\.attn_v\.weight$=q5_k
+^blk\.34\.ffn_down\.weight$=q5_k
+^blk\.34\.ffn_gate\.weight$=iq3_s
+^blk\.34\.ffn_up\.weight$=q5_k
+^blk\.35\.attn_k\.weight$=q5_k
+^blk\.35\.attn_output\.weight$=q5_k
+^blk\.35\.attn_q\.weight$=iq3_s
+^blk\.35\.attn_v\.weight$=q6_k
+^blk\.35\.ffn_down\.weight$=q5_k
+^blk\.35\.ffn_gate\.weight$=q5_k
+^blk\.35\.ffn_up\.weight$=q5_k

layouts/AesCode-8B-IQ3_XS.layout.json

@@ -0,0 +1,84 @@
+{
+  "model": "/media/m2teamgroup/models/AesCode-8B-GGUF/AesCode-8B-bf16.gguf",
+  "imatrix": "/media/m2teamgroup/models/AesCode-8B-GGUF/AesCode-8B-imatrix.gguf",
+  "ftype": "IQ3_XS",
+  "base_type": "iq3_s",
+  "family": "iq",
+  "recipe": "ident",
+  "variant": "shr0.90",
+  "auto_crush": false,
+  "rung_mode": "share",
+  "tier": "S",
+  "rung_share": 0.9,
+  "crush_pass_bytes": null,
+  "embd_shares": {
+    "output.weight": {
+      "share": 0.1353,
+      "type": "q5_k"
+    },
+    "token_embd.weight": {
+      "share": 0.0709,
+      "type": "q4_k"
+    }
+  },
+  "arch": "dense",
+  "tied": false,
+  "n_layer": 36,
+  "body_rule": true,
+  "pins": {},
+  "shape_pins": {},
+  "fixed_pins": {},
+  "embeddings": {
+    "output.weight": "q5_k",
+    "token_embd.weight": "q4_k"
+  },
+  "bytes": {
+    "heuristic_body+embd": 3773038592,
+    "map_body+embd": 3879239680,
+    "ratio": 1.0281,
+    "note": "quantizable, unpinned tensors only; pinned/unquantized tensors are identical in both"
+  },
+  "body_base_share": 0.9,
+  "body_bpw": 3.572,
+  "body_bpw_heuristic": 3.45,
+  "total_params": 8190735360,
+  "file_bytes_pred": 3880472576,
+  "file_bpw_pred": 3.79,
+  "histogram": {
+    "attn_k": {
+      "q5_k": 27,
+      "iq3_s": 9
+    },
+    "attn_output": {
+      "iq3_s": 28,
+      "q5_k": 8
+    },
+    "attn_q": {
+      "iq3_s": 36
+    },
+    "attn_v": {
+      "q5_k": 34,
+      "q6_k": 2
+    },
+    "ffn_down": {
+      "iq3_s": 36
+    },
+    "ffn_gate": {
+      "iq3_s": 36
+    },
+    "ffn_up": {
+      "iq3_s": 35,
+      "q5_k": 1
+    }
+  },
+  "prior": "layout/prior.json",
+  "pin_rule": "floor",
+  "embd_cap": "heur",
+  "generator": "auto_quant_v2.layout",
+  "generator_key": "3bf8b43e20a68f0b",
+  "quant_type": "IQ3_XS",
+  "quantize_ftype": "IQ3_XS",
+  "llama_cpp_version": "b11490",
+  "unknown_kinds": {},
+  "commit": "5d550ef5561a30307f1d436db2c6835b75395e77"
+}

layouts/AesCode-8B-IQ3_XS.tensor-types.txt

@@ -0,0 +1,254 @@
+^output\.weight$=q5_k
+^token_embd\.weight$=q4_k
+^blk\.0\.attn_k\.weight$=q5_k
+^blk\.0\.attn_output\.weight$=iq3_s
+^blk\.0\.attn_q\.weight$=iq3_s
+^blk\.0\.attn_v\.weight$=q5_k
+^blk\.0\.ffn_down\.weight$=iq3_s
+^blk\.0\.ffn_gate\.weight$=iq3_s
+^blk\.0\.ffn_up\.weight$=iq3_s
+^blk\.1\.attn_k\.weight$=q5_k
+^blk\.1\.attn_output\.weight$=q5_k
+^blk\.1\.attn_q\.weight$=iq3_s
+^blk\.1\.attn_v\.weight$=q6_k
+^blk\.1\.ffn_down\.weight$=iq3_s
+^blk\.1\.ffn_gate\.weight$=iq3_s
+^blk\.1\.ffn_up\.weight$=iq3_s
+^blk\.2\.attn_k\.weight$=q5_k
+^blk\.2\.attn_output\.weight$=q5_k
+^blk\.2\.attn_q\.weight$=iq3_s
+^blk\.2\.attn_v\.weight$=q6_k
+^blk\.2\.ffn_down\.weight$=iq3_s
+^blk\.2\.ffn_gate\.weight$=iq3_s
+^blk\.2\.ffn_up\.weight$=iq3_s
+^blk\.3\.attn_k\.weight$=q5_k
+^blk\.3\.attn_output\.weight$=q5_k
+^blk\.3\.attn_q\.weight$=iq3_s
+^blk\.3\.attn_v\.weight$=q5_k
+^blk\.3\.ffn_down\.weight$=iq3_s
+^blk\.3\.ffn_gate\.weight$=iq3_s
+^blk\.3\.ffn_up\.weight$=iq3_s
+^blk\.4\.attn_k\.weight$=q5_k
+^blk\.4\.attn_output\.weight$=q5_k
+^blk\.4\.attn_q\.weight$=iq3_s
+^blk\.4\.attn_v\.weight$=q5_k
+^blk\.4\.ffn_down\.weight$=iq3_s
+^blk\.4\.ffn_gate\.weight$=iq3_s
+^blk\.4\.ffn_up\.weight$=iq3_s
+^blk\.5\.attn_k\.weight$=q5_k
+^blk\.5\.attn_output\.weight$=q5_k
+^blk\.5\.attn_q\.weight$=iq3_s
+^blk\.5\.attn_v\.weight$=q5_k
+^blk\.5\.ffn_down\.weight$=iq3_s
+^blk\.5\.ffn_gate\.weight$=iq3_s
+^blk\.5\.ffn_up\.weight$=iq3_s
+^blk\.6\.attn_k\.weight$=q5_k
+^blk\.6\.attn_output\.weight$=q5_k
+^blk\.6\.attn_q\.weight$=iq3_s
+^blk\.6\.attn_v\.weight$=q5_k
+^blk\.6\.ffn_down\.weight$=iq3_s
+^blk\.6\.ffn_gate\.weight$=iq3_s
+^blk\.6\.ffn_up\.weight$=iq3_s
+^blk\.7\.attn_k\.weight$=q5_k
+^blk\.7\.attn_output\.weight$=q5_k
+^blk\.7\.attn_q\.weight$=iq3_s
+^blk\.7\.attn_v\.weight$=q5_k
+^blk\.7\.ffn_down\.weight$=iq3_s
+^blk\.7\.ffn_gate\.weight$=iq3_s
+^blk\.7\.ffn_up\.weight$=iq3_s
+^blk\.8\.attn_k\.weight$=q5_k
+^blk\.8\.attn_output\.weight$=q5_k
+^blk\.8\.attn_q\.weight$=iq3_s
+^blk\.8\.attn_v\.weight$=q5_k
+^blk\.8\.ffn_down\.weight$=iq3_s
+^blk\.8\.ffn_gate\.weight$=iq3_s
+^blk\.8\.ffn_up\.weight$=iq3_s
+^blk\.9\.attn_k\.weight$=q5_k
+^blk\.9\.attn_output\.weight$=iq3_s
+^blk\.9\.attn_q\.weight$=iq3_s
+^blk\.9\.attn_v\.weight$=q5_k
+^blk\.9\.ffn_down\.weight$=iq3_s
+^blk\.9\.ffn_gate\.weight$=iq3_s
+^blk\.9\.ffn_up\.weight$=iq3_s
+^blk\.10\.attn_k\.weight$=q5_k
+^blk\.10\.attn_output\.weight$=iq3_s
+^blk\.10\.attn_q\.weight$=iq3_s
+^blk\.10\.attn_v\.weight$=q5_k
+^blk\.10\.ffn_down\.weight$=iq3_s
+^blk\.10\.ffn_gate\.weight$=iq3_s
+^blk\.10\.ffn_up\.weight$=iq3_s
+^blk\.11\.attn_k\.weight$=q5_k
+^blk\.11\.attn_output\.weight$=iq3_s
+^blk\.11\.attn_q\.weight$=iq3_s
+^blk\.11\.attn_v\.weight$=q5_k
+^blk\.11\.ffn_down\.weight$=iq3_s
+^blk\.11\.ffn_gate\.weight$=iq3_s
+^blk\.11\.ffn_up\.weight$=iq3_s
+^blk\.12\.attn_k\.weight$=q5_k
+^blk\.12\.attn_output\.weight$=iq3_s
+^blk\.12\.attn_q\.weight$=iq3_s
+^blk\.12\.attn_v\.weight$=q5_k
+^blk\.12\.ffn_down\.weight$=iq3_s
+^blk\.12\.ffn_gate\.weight$=iq3_s
+^blk\.12\.ffn_up\.weight$=iq3_s
+^blk\.13\.attn_k\.weight$=q5_k
+^blk\.13\.attn_output\.weight$=iq3_s
+^blk\.13\.attn_q\.weight$=iq3_s
+^blk\.13\.attn_v\.weight$=q5_k
+^blk\.13\.ffn_down\.weight$=iq3_s
+^blk\.13\.ffn_gate\.weight$=iq3_s
+^blk\.13\.ffn_up\.weight$=iq3_s
+^blk\.14\.attn_k\.weight$=q5_k
+^blk\.14\.attn_output\.weight$=iq3_s
+^blk\.14\.attn_q\.weight$=iq3_s
+^blk\.14\.attn_v\.weight$=q5_k
+^blk\.14\.ffn_down\.weight$=iq3_s
+^blk\.14\.ffn_gate\.weight$=iq3_s
+^blk\.14\.ffn_up\.weight$=iq3_s
+^blk\.15\.attn_k\.weight$=q5_k
+^blk\.15\.attn_output\.weight$=iq3_s
+^blk\.15\.attn_q\.weight$=iq3_s
+^blk\.15\.attn_v\.weight$=q5_k
+^blk\.15\.ffn_down\.weight$=iq3_s
+^blk\.15\.ffn_gate\.weight$=iq3_s
+^blk\.15\.ffn_up\.weight$=iq3_s
+^blk\.16\.attn_k\.weight$=q5_k
+^blk\.16\.attn_output\.weight$=iq3_s
+^blk\.16\.attn_q\.weight$=iq3_s
+^blk\.16\.attn_v\.weight$=q5_k
+^blk\.16\.ffn_down\.weight$=iq3_s
+^blk\.16\.ffn_gate\.weight$=iq3_s
+^blk\.16\.ffn_up\.weight$=iq3_s
+^blk\.17\.attn_k\.weight$=q5_k
+^blk\.17\.attn_output\.weight$=iq3_s
+^blk\.17\.attn_q\.weight$=iq3_s
+^blk\.17\.attn_v\.weight$=q5_k
+^blk\.17\.ffn_down\.weight$=iq3_s
+^blk\.17\.ffn_gate\.weight$=iq3_s
+^blk\.17\.ffn_up\.weight$=iq3_s
+^blk\.18\.attn_k\.weight$=q5_k
+^blk\.18\.attn_output\.weight$=iq3_s
+^blk\.18\.attn_q\.weight$=iq3_s
+^blk\.18\.attn_v\.weight$=q5_k
+^blk\.18\.ffn_down\.weight$=iq3_s
+^blk\.18\.ffn_gate\.weight$=iq3_s
+^blk\.18\.ffn_up\.weight$=iq3_s
+^blk\.19\.attn_k\.weight$=q5_k
+^blk\.19\.attn_output\.weight$=iq3_s
+^blk\.19\.attn_q\.weight$=iq3_s
+^blk\.19\.attn_v\.weight$=q5_k
+^blk\.19\.ffn_down\.weight$=iq3_s
+^blk\.19\.ffn_gate\.weight$=iq3_s
+^blk\.19\.ffn_up\.weight$=iq3_s
+^blk\.20\.attn_k\.weight$=q5_k
+^blk\.20\.attn_output\.weight$=iq3_s
+^blk\.20\.attn_q\.weight$=iq3_s
+^blk\.20\.attn_v\.weight$=q5_k
+^blk\.20\.ffn_down\.weight$=iq3_s
+^blk\.20\.ffn_gate\.weight$=iq3_s
+^blk\.20\.ffn_up\.weight$=iq3_s
+^blk\.21\.attn_k\.weight$=q5_k
+^blk\.21\.attn_output\.weight$=iq3_s
+^blk\.21\.attn_q\.weight$=iq3_s
+^blk\.21\.attn_v\.weight$=q5_k
+^blk\.21\.ffn_down\.weight$=iq3_s
+^blk\.21\.ffn_gate\.weight$=iq3_s
+^blk\.21\.ffn_up\.weight$=iq3_s
+^blk\.22\.attn_k\.weight$=q5_k
+^blk\.22\.attn_output\.weight$=iq3_s
+^blk\.22\.attn_q\.weight$=iq3_s
+^blk\.22\.attn_v\.weight$=q5_k
+^blk\.22\.ffn_down\.weight$=iq3_s
+^blk\.22\.ffn_gate\.weight$=iq3_s
+^blk\.22\.ffn_up\.weight$=iq3_s
+^blk\.23\.attn_k\.weight$=q5_k
+^blk\.23\.attn_output\.weight$=iq3_s
+^blk\.23\.attn_q\.weight$=iq3_s
+^blk\.23\.attn_v\.weight$=q5_k
+^blk\.23\.ffn_down\.weight$=iq3_s
+^blk\.23\.ffn_gate\.weight$=iq3_s
+^blk\.23\.ffn_up\.weight$=iq3_s
+^blk\.24\.attn_k\.weight$=q5_k
+^blk\.24\.attn_output\.weight$=iq3_s
+^blk\.24\.attn_q\.weight$=iq3_s
+^blk\.24\.attn_v\.weight$=q5_k
+^blk\.24\.ffn_down\.weight$=iq3_s
+^blk\.24\.ffn_gate\.weight$=iq3_s
+^blk\.24\.ffn_up\.weight$=iq3_s
+^blk\.25\.attn_k\.weight$=q5_k
+^blk\.25\.attn_output\.weight$=iq3_s
+^blk\.25\.attn_q\.weight$=iq3_s
+^blk\.25\.attn_v\.weight$=q5_k
+^blk\.25\.ffn_down\.weight$=iq3_s
+^blk\.25\.ffn_gate\.weight$=iq3_s
+^blk\.25\.ffn_up\.weight$=iq3_s
+^blk\.26\.attn_k\.weight$=q5_k
+^blk\.26\.attn_output\.weight$=iq3_s
+^blk\.26\.attn_q\.weight$=iq3_s
+^blk\.26\.attn_v\.weight$=q5_k
+^blk\.26\.ffn_down\.weight$=iq3_s
+^blk\.26\.ffn_gate\.weight$=iq3_s
+^blk\.26\.ffn_up\.weight$=iq3_s
+^blk\.27\.attn_k\.weight$=iq3_s
+^blk\.27\.attn_output\.weight$=iq3_s
+^blk\.27\.attn_q\.weight$=iq3_s
+^blk\.27\.attn_v\.weight$=q5_k
+^blk\.27\.ffn_down\.weight$=iq3_s
+^blk\.27\.ffn_gate\.weight$=iq3_s
+^blk\.27\.ffn_up\.weight$=iq3_s
+^blk\.28\.attn_k\.weight$=iq3_s
+^blk\.28\.attn_output\.weight$=iq3_s
+^blk\.28\.attn_q\.weight$=iq3_s
+^blk\.28\.attn_v\.weight$=q5_k
+^blk\.28\.ffn_down\.weight$=iq3_s
+^blk\.28\.ffn_gate\.weight$=iq3_s
+^blk\.28\.ffn_up\.weight$=iq3_s
+^blk\.29\.attn_k\.weight$=iq3_s
+^blk\.29\.attn_output\.weight$=iq3_s
+^blk\.29\.attn_q\.weight$=iq3_s
+^blk\.29\.attn_v\.weight$=q5_k
+^blk\.29\.ffn_down\.weight$=iq3_s
+^blk\.29\.ffn_gate\.weight$=iq3_s
+^blk\.29\.ffn_up\.weight$=iq3_s
+^blk\.30\.attn_k\.weight$=iq3_s
+^blk\.30\.attn_output\.weight$=iq3_s
+^blk\.30\.attn_q\.weight$=iq3_s
+^blk\.30\.attn_v\.weight$=q5_k
+^blk\.30\.ffn_down\.weight$=iq3_s
+^blk\.30\.ffn_gate\.weight$=iq3_s
+^blk\.30\.ffn_up\.weight$=iq3_s
+^blk\.31\.attn_k\.weight$=iq3_s
+^blk\.31\.attn_output\.weight$=iq3_s
+^blk\.31\.attn_q\.weight$=iq3_s
+^blk\.31\.attn_v\.weight$=q5_k
+^blk\.31\.ffn_down\.weight$=iq3_s
+^blk\.31\.ffn_gate\.weight$=iq3_s
+^blk\.31\.ffn_up\.weight$=iq3_s
+^blk\.32\.attn_k\.weight$=iq3_s
+^blk\.32\.attn_output\.weight$=iq3_s
+^blk\.32\.attn_q\.weight$=iq3_s
+^blk\.32\.attn_v\.weight$=q5_k
+^blk\.32\.ffn_down\.weight$=iq3_s
+^blk\.32\.ffn_gate\.weight$=iq3_s
+^blk\.32\.ffn_up\.weight$=iq3_s
+^blk\.33\.attn_k\.weight$=iq3_s
+^blk\.33\.attn_output\.weight$=iq3_s
+^blk\.33\.attn_q\.weight$=iq3_s
+^blk\.33\.attn_v\.weight$=q5_k
+^blk\.33\.ffn_down\.weight$=iq3_s
+^blk\.33\.ffn_gate\.weight$=iq3_s
+^blk\.33\.ffn_up\.weight$=iq3_s
+^blk\.34\.attn_k\.weight$=iq3_s
+^blk\.34\.attn_output\.weight$=iq3_s
+^blk\.34\.attn_q\.weight$=iq3_s
+^blk\.34\.attn_v\.weight$=q5_k
+^blk\.34\.ffn_down\.weight$=iq3_s
+^blk\.34\.ffn_gate\.weight$=iq3_s
+^blk\.34\.ffn_up\.weight$=iq3_s
+^blk\.35\.attn_k\.weight$=iq3_s
+^blk\.35\.attn_output\.weight$=iq3_s
+^blk\.35\.attn_q\.weight$=iq3_s
+^blk\.35\.attn_v\.weight$=q5_k
+^blk\.35\.ffn_down\.weight$=iq3_s
+^blk\.35\.ffn_gate\.weight$=iq3_s
+^blk\.35\.ffn_up\.weight$=q5_k

layouts/AesCode-8B-IQ3_XXS.layout.json

@@ -0,0 +1,87 @@
+{
+  "model": "/media/m2teamgroup/models/AesCode-8B-GGUF/AesCode-8B-bf16.gguf",
+  "imatrix": "/media/m2teamgroup/models/AesCode-8B-GGUF/AesCode-8B-imatrix.gguf",
+  "ftype": "IQ3_XXS",
+  "base_type": "iq3_xxs",
+  "family": "iq",
+  "recipe": "ident",
+  "variant": "shr0.70",
+  "auto_crush": false,
+  "rung_mode": "share",
+  "tier": "M",
+  "rung_share": 0.7,
+  "crush_pass_bytes": null,
+  "embd_shares": {
+    "output.weight": {
+      "share": 0.1208,
+      "type": "q5_k"
+    },
+    "token_embd.weight": {
+      "share": 0.0755,
+      "type": "q4_k"
+    }
+  },
+  "arch": "dense",
+  "tied": false,
+  "n_layer": 36,
+  "body_rule": true,
+  "pins": {},
+  "shape_pins": {},
+  "fixed_pins": {},
+  "embeddings": {
+    "output.weight": "q5_k",
+    "token_embd.weight": "q4_k"
+  },
+  "bytes": {
+    "heuristic_body+embd": 3541749760,
+    "map_body+embd": 3677192192,
+    "ratio": 1.0382,
+    "note": "quantizable, unpinned tensors only; pinned/unquantized tensors are identical in both"
+  },
+  "body_base_share": 0.7,
+  "body_bpw": 3.339,
+  "body_bpw_heuristic": 3.279,
+  "total_params": 8190735360,
+  "file_bytes_pred": 3678425088,
+  "file_bpw_pred": 3.593,
+  "histogram": {
+    "attn_k": {
+      "q4_k": 18,
+      "iq3_s": 9,
+      "iq4_xs": 8,
+      "q5_k": 1
+    },
+    "attn_output": {
+      "iq3_s": 18,
+      "iq3_xxs": 10,
+      "q4_k": 8
+    },
+    "attn_q": {
+      "iq3_xxs": 36
+    },
+    "attn_v": {
+      "q5_k": 36
+    },
+    "ffn_down": {
+      "iq3_xxs": 27,
+      "iq4_xs": 9
+    },
+    "ffn_gate": {
+      "iq3_xxs": 36
+    },
+    "ffn_up": {
+      "iq3_xxs": 27,
+      "iq4_xs": 9
+    }
+  },
+  "prior": "layout/prior.json",
+  "pin_rule": "floor",
+  "embd_cap": "heur",
+  "generator": "auto_quant_v2.layout",
+  "generator_key": "3bf8b43e20a68f0b",
+  "quant_type": "IQ3_XXS",
+  "quantize_ftype": "IQ3_XXS",
+  "llama_cpp_version": "b11490",
+  "unknown_kinds": {},
+  "commit": "5d550ef5561a30307f1d436db2c6835b75395e77"
+}

layouts/AesCode-8B-IQ3_XXS.tensor-types.txt

@@ -0,0 +1,254 @@
+^output\.weight$=q5_k
+^token_embd\.weight$=q4_k
+^blk\.0\.attn_k\.weight$=q5_k
+^blk\.0\.attn_output\.weight$=iq3_xxs
+^blk\.0\.attn_q\.weight$=iq3_xxs
+^blk\.0\.attn_v\.weight$=q5_k
+^blk\.0\.ffn_down\.weight$=iq3_xxs
+^blk\.0\.ffn_gate\.weight$=iq3_xxs
+^blk\.0\.ffn_up\.weight$=iq3_xxs
+^blk\.1\.attn_k\.weight$=iq4_xs
+^blk\.1\.attn_output\.weight$=q4_k
+^blk\.1\.attn_q\.weight$=iq3_xxs
+^blk\.1\.attn_v\.weight$=q5_k
+^blk\.1\.ffn_down\.weight$=iq4_xs
+^blk\.1\.ffn_gate\.weight$=iq3_xxs
+^blk\.1\.ffn_up\.weight$=iq4_xs
+^blk\.2\.attn_k\.weight$=iq4_xs
+^blk\.2\.attn_output\.weight$=q4_k
+^blk\.2\.attn_q\.weight$=iq3_xxs
+^blk\.2\.attn_v\.weight$=q5_k
+^blk\.2\.ffn_down\.weight$=iq4_xs
+^blk\.2\.ffn_gate\.weight$=iq3_xxs
+^blk\.2\.ffn_up\.weight$=iq4_xs
+^blk\.3\.attn_k\.weight$=iq4_xs
+^blk\.3\.attn_output\.weight$=q4_k
+^blk\.3\.attn_q\.weight$=iq3_xxs
+^blk\.3\.attn_v\.weight$=q5_k
+^blk\.3\.ffn_down\.weight$=iq4_xs
+^blk\.3\.ffn_gate\.weight$=iq3_xxs
+^blk\.3\.ffn_up\.weight$=iq4_xs
+^blk\.4\.attn_k\.weight$=iq4_xs
+^blk\.4\.attn_output\.weight$=q4_k
+^blk\.4\.attn_q\.weight$=iq3_xxs
+^blk\.4\.attn_v\.weight$=q5_k
+^blk\.4\.ffn_down\.weight$=iq4_xs
+^blk\.4\.ffn_gate\.weight$=iq3_xxs
+^blk\.4\.ffn_up\.weight$=iq4_xs
+^blk\.5\.attn_k\.weight$=iq4_xs
+^blk\.5\.attn_output\.weight$=q4_k
+^blk\.5\.attn_q\.weight$=iq3_xxs
+^blk\.5\.attn_v\.weight$=q5_k
+^blk\.5\.ffn_down\.weight$=iq4_xs
+^blk\.5\.ffn_gate\.weight$=iq3_xxs
+^blk\.5\.ffn_up\.weight$=iq4_xs
+^blk\.6\.attn_k\.weight$=iq4_xs
+^blk\.6\.attn_output\.weight$=q4_k
+^blk\.6\.attn_q\.weight$=iq3_xxs
+^blk\.6\.attn_v\.weight$=q5_k
+^blk\.6\.ffn_down\.weight$=iq4_xs
+^blk\.6\.ffn_gate\.weight$=iq3_xxs
+^blk\.6\.ffn_up\.weight$=iq4_xs
+^blk\.7\.attn_k\.weight$=iq4_xs
+^blk\.7\.attn_output\.weight$=q4_k
+^blk\.7\.attn_q\.weight$=iq3_xxs
+^blk\.7\.attn_v\.weight$=q5_k
+^blk\.7\.ffn_down\.weight$=iq4_xs
+^blk\.7\.ffn_gate\.weight$=iq3_xxs
+^blk\.7\.ffn_up\.weight$=iq4_xs
+^blk\.8\.attn_k\.weight$=iq4_xs
+^blk\.8\.attn_output\.weight$=q4_k
+^blk\.8\.attn_q\.weight$=iq3_xxs
+^blk\.8\.attn_v\.weight$=q5_k
+^blk\.8\.ffn_down\.weight$=iq4_xs
+^blk\.8\.ffn_gate\.weight$=iq3_xxs
+^blk\.8\.ffn_up\.weight$=iq4_xs
+^blk\.9\.attn_k\.weight$=q4_k
+^blk\.9\.attn_output\.weight$=iq3_s
+^blk\.9\.attn_q\.weight$=iq3_xxs
+^blk\.9\.attn_v\.weight$=q5_k
+^blk\.9\.ffn_down\.weight$=iq3_xxs
+^blk\.9\.ffn_gate\.weight$=iq3_xxs
+^blk\.9\.ffn_up\.weight$=iq3_xxs
+^blk\.10\.attn_k\.weight$=q4_k
+^blk\.10\.attn_output\.weight$=iq3_s
+^blk\.10\.attn_q\.weight$=iq3_xxs
+^blk\.10\.attn_v\.weight$=q5_k
+^blk\.10\.ffn_down\.weight$=iq3_xxs
+^blk\.10\.ffn_gate\.weight$=iq3_xxs
+^blk\.10\.ffn_up\.weight$=iq3_xxs
+^blk\.11\.attn_k\.weight$=q4_k
+^blk\.11\.attn_output\.weight$=iq3_s
+^blk\.11\.attn_q\.weight$=iq3_xxs
+^blk\.11\.attn_v\.weight$=q5_k
+^blk\.11\.ffn_down\.weight$=iq3_xxs
+^blk\.11\.ffn_gate\.weight$=iq3_xxs
+^blk\.11\.ffn_up\.weight$=iq3_xxs
+^blk\.12\.attn_k\.weight$=q4_k
+^blk\.12\.attn_output\.weight$=iq3_s
+^blk\.12\.attn_q\.weight$=iq3_xxs
+^blk\.12\.attn_v\.weight$=q5_k
+^blk\.12\.ffn_down\.weight$=iq3_xxs
+^blk\.12\.ffn_gate\.weight$=iq3_xxs
+^blk\.12\.ffn_up\.weight$=iq3_xxs
+^blk\.13\.attn_k\.weight$=q4_k
+^blk\.13\.attn_output\.weight$=iq3_s
+^blk\.13\.attn_q\.weight$=iq3_xxs
+^blk\.13\.attn_v\.weight$=q5_k
+^blk\.13\.ffn_down\.weight$=iq3_xxs
+^blk\.13\.ffn_gate\.weight$=iq3_xxs
+^blk\.13\.ffn_up\.weight$=iq3_xxs
+^blk\.14\.attn_k\.weight$=q4_k
+^blk\.14\.attn_output\.weight$=iq3_s
+^blk\.14\.attn_q\.weight$=iq3_xxs
+^blk\.14\.attn_v\.weight$=q5_k
+^blk\.14\.ffn_down\.weight$=iq3_xxs
+^blk\.14\.ffn_gate\.weight$=iq3_xxs
+^blk\.14\.ffn_up\.weight$=iq3_xxs
+^blk\.15\.attn_k\.weight$=q4_k
+^blk\.15\.attn_output\.weight$=iq3_s
+^blk\.15\.attn_q\.weight$=iq3_xxs
+^blk\.15\.attn_v\.weight$=q5_k
+^blk\.15\.ffn_down\.weight$=iq3_xxs
+^blk\.15\.ffn_gate\.weight$=iq3_xxs
+^blk\.15\.ffn_up\.weight$=iq3_xxs
+^blk\.16\.attn_k\.weight$=q4_k
+^blk\.16\.attn_output\.weight$=iq3_s
+^blk\.16\.attn_q\.weight$=iq3_xxs
+^blk\.16\.attn_v\.weight$=q5_k
+^blk\.16\.ffn_down\.weight$=iq3_xxs
+^blk\.16\.ffn_gate\.weight$=iq3_xxs
+^blk\.16\.ffn_up\.weight$=iq3_xxs
+^blk\.17\.attn_k\.weight$=q4_k
+^blk\.17\.attn_output\.weight$=iq3_s
+^blk\.17\.attn_q\.weight$=iq3_xxs
+^blk\.17\.attn_v\.weight$=q5_k
+^blk\.17\.ffn_down\.weight$=iq3_xxs
+^blk\.17\.ffn_gate\.weight$=iq3_xxs
+^blk\.17\.ffn_up\.weight$=iq3_xxs
+^blk\.18\.attn_k\.weight$=q4_k
+^blk\.18\.attn_output\.weight$=iq3_s
+^blk\.18\.attn_q\.weight$=iq3_xxs
+^blk\.18\.attn_v\.weight$=q5_k
+^blk\.18\.ffn_down\.weight$=iq3_xxs
+^blk\.18\.ffn_gate\.weight$=iq3_xxs
+^blk\.18\.ffn_up\.weight$=iq3_xxs
+^blk\.19\.attn_k\.weight$=q4_k
+^blk\.19\.attn_output\.weight$=iq3_s
+^blk\.19\.attn_q\.weight$=iq3_xxs
+^blk\.19\.attn_v\.weight$=q5_k
+^blk\.19\.ffn_down\.weight$=iq3_xxs
+^blk\.19\.ffn_gate\.weight$=iq3_xxs
+^blk\.19\.ffn_up\.weight$=iq3_xxs
+^blk\.20\.attn_k\.weight$=q4_k
+^blk\.20\.attn_output\.weight$=iq3_s
+^blk\.20\.attn_q\.weight$=iq3_xxs
+^blk\.20\.attn_v\.weight$=q5_k
+^blk\.20\.ffn_down\.weight$=iq3_xxs
+^blk\.20\.ffn_gate\.weight$=iq3_xxs
+^blk\.20\.ffn_up\.weight$=iq3_xxs
+^blk\.21\.attn_k\.weight$=q4_k
+^blk\.21\.attn_output\.weight$=iq3_s
+^blk\.21\.attn_q\.weight$=iq3_xxs
+^blk\.21\.attn_v\.weight$=q5_k
+^blk\.21\.ffn_down\.weight$=iq3_xxs
+^blk\.21\.ffn_gate\.weight$=iq3_xxs
+^blk\.21\.ffn_up\.weight$=iq3_xxs
+^blk\.22\.attn_k\.weight$=q4_k
+^blk\.22\.attn_output\.weight$=iq3_s
+^blk\.22\.attn_q\.weight$=iq3_xxs
+^blk\.22\.attn_v\.weight$=q5_k
+^blk\.22\.ffn_down\.weight$=iq3_xxs
+^blk\.22\.ffn_gate\.weight$=iq3_xxs
+^blk\.22\.ffn_up\.weight$=iq3_xxs
+^blk\.23\.attn_k\.weight$=q4_k
+^blk\.23\.attn_output\.weight$=iq3_s
+^blk\.23\.attn_q\.weight$=iq3_xxs
+^blk\.23\.attn_v\.weight$=q5_k
+^blk\.23\.ffn_down\.weight$=iq3_xxs
+^blk\.23\.ffn_gate\.weight$=iq3_xxs
+^blk\.23\.ffn_up\.weight$=iq3_xxs
+^blk\.24\.attn_k\.weight$=q4_k
+^blk\.24\.attn_output\.weight$=iq3_s
+^blk\.24\.attn_q\.weight$=iq3_xxs
+^blk\.24\.attn_v\.weight$=q5_k
+^blk\.24\.ffn_down\.weight$=iq3_xxs
+^blk\.24\.ffn_gate\.weight$=iq3_xxs
+^blk\.24\.ffn_up\.weight$=iq3_xxs
+^blk\.25\.attn_k\.weight$=q4_k
+^blk\.25\.attn_output\.weight$=iq3_s
+^blk\.25\.attn_q\.weight$=iq3_xxs
+^blk\.25\.attn_v\.weight$=q5_k
+^blk\.25\.ffn_down\.weight$=iq3_xxs
+^blk\.25\.ffn_gate\.weight$=iq3_xxs
+^blk\.25\.ffn_up\.weight$=iq3_xxs
+^blk\.26\.attn_k\.weight$=q4_k
+^blk\.26\.attn_output\.weight$=iq3_s
+^blk\.26\.attn_q\.weight$=iq3_xxs
+^blk\.26\.attn_v\.weight$=q5_k
+^blk\.26\.ffn_down\.weight$=iq3_xxs
+^blk\.26\.ffn_gate\.weight$=iq3_xxs
+^blk\.26\.ffn_up\.weight$=iq3_xxs
+^blk\.27\.attn_k\.weight$=iq3_s
+^blk\.27\.attn_output\.weight$=iq3_xxs
+^blk\.27\.attn_q\.weight$=iq3_xxs
+^blk\.27\.attn_v\.weight$=q5_k
+^blk\.27\.ffn_down\.weight$=iq3_xxs
+^blk\.27\.ffn_gate\.weight$=iq3_xxs
+^blk\.27\.ffn_up\.weight$=iq3_xxs
+^blk\.28\.attn_k\.weight$=iq3_s
+^blk\.28\.attn_output\.weight$=iq3_xxs
+^blk\.28\.attn_q\.weight$=iq3_xxs
+^blk\.28\.attn_v\.weight$=q5_k
+^blk\.28\.ffn_down\.weight$=iq3_xxs
+^blk\.28\.ffn_gate\.weight$=iq3_xxs
+^blk\.28\.ffn_up\.weight$=iq3_xxs
+^blk\.29\.attn_k\.weight$=iq3_s
+^blk\.29\.attn_output\.weight$=iq3_xxs
+^blk\.29\.attn_q\.weight$=iq3_xxs
+^blk\.29\.attn_v\.weight$=q5_k
+^blk\.29\.ffn_down\.weight$=iq3_xxs
+^blk\.29\.ffn_gate\.weight$=iq3_xxs
+^blk\.29\.ffn_up\.weight$=iq3_xxs
+^blk\.30\.attn_k\.weight$=iq3_s
+^blk\.30\.attn_output\.weight$=iq3_xxs
+^blk\.30\.attn_q\.weight$=iq3_xxs
+^blk\.30\.attn_v\.weight$=q5_k
+^blk\.30\.ffn_down\.weight$=iq3_xxs
+^blk\.30\.ffn_gate\.weight$=iq3_xxs
+^blk\.30\.ffn_up\.weight$=iq3_xxs
+^blk\.31\.attn_k\.weight$=iq3_s
+^blk\.31\.attn_output\.weight$=iq3_xxs
+^blk\.31\.attn_q\.weight$=iq3_xxs
+^blk\.31\.attn_v\.weight$=q5_k
+^blk\.31\.ffn_down\.weight$=iq3_xxs
+^blk\.31\.ffn_gate\.weight$=iq3_xxs
+^blk\.31\.ffn_up\.weight$=iq3_xxs
+^blk\.32\.attn_k\.weight$=iq3_s
+^blk\.32\.attn_output\.weight$=iq3_xxs
+^blk\.32\.attn_q\.weight$=iq3_xxs
+^blk\.32\.attn_v\.weight$=q5_k
+^blk\.32\.ffn_down\.weight$=iq3_xxs
+^blk\.32\.ffn_gate\.weight$=iq3_xxs
+^blk\.32\.ffn_up\.weight$=iq3_xxs
+^blk\.33\.attn_k\.weight$=iq3_s
+^blk\.33\.attn_output\.weight$=iq3_xxs
+^blk\.33\.attn_q\.weight$=iq3_xxs
+^blk\.33\.attn_v\.weight$=q5_k
+^blk\.33\.ffn_down\.weight$=iq3_xxs
+^blk\.33\.ffn_gate\.weight$=iq3_xxs
+^blk\.33\.ffn_up\.weight$=iq3_xxs
+^blk\.34\.attn_k\.weight$=iq3_s
+^blk\.34\.attn_output\.weight$=iq3_xxs
+^blk\.34\.attn_q\.weight$=iq3_xxs
+^blk\.34\.attn_v\.weight$=q5_k
+^blk\.34\.ffn_down\.weight$=iq3_xxs
+^blk\.34\.ffn_gate\.weight$=iq3_xxs
+^blk\.34\.ffn_up\.weight$=iq3_xxs
+^blk\.35\.attn_k\.weight$=iq3_s
+^blk\.35\.attn_output\.weight$=iq3_xxs
+^blk\.35\.attn_q\.weight$=iq3_xxs
+^blk\.35\.attn_v\.weight$=q5_k
+^blk\.35\.ffn_down\.weight$=iq4_xs
+^blk\.35\.ffn_gate\.weight$=iq3_xxs
+^blk\.35\.ffn_up\.weight$=iq4_xs

layouts/AesCode-8B-IQ4_NL.layout.json

@@ -0,0 +1,87 @@
+{
+  "model": "/media/m2teamgroup/models/AesCode-8B-GGUF/AesCode-8B-bf16.gguf",
+  "imatrix": "/media/m2teamgroup/models/AesCode-8B-GGUF/AesCode-8B-imatrix.gguf",
+  "ftype": "IQ4_NL",
+  "base_type": "iq4_nl",
+  "family": "iq",
+  "recipe": "ident",
+  "variant": "shr0.70",
+  "auto_crush": false,
+  "rung_mode": "share",
+  "tier": "M",
+  "rung_share": 0.7,
+  "crush_pass_bytes": null,
+  "embd_shares": {
+    "output.weight": {
+      "share": 0.1045,
+      "type": "q6_k"
+    },
+    "token_embd.weight": {
+      "share": 0.0717,
+      "type": "iq4_nl"
+    }
+  },
+  "arch": "dense",
+  "tied": false,
+  "n_layer": 36,
+  "body_rule": true,
+  "pins": {},
+  "shape_pins": {},
+  "fixed_pins": {},
+  "embeddings": {
+    "output.weight": "q6_k",
+    "token_embd.weight": "iq4_nl"
+  },
+  "bytes": {
+    "heuristic_body+embd": 4884344832,
+    "map_body+embd": 5171261440,
+    "ratio": 1.0587,
+    "note": "quantizable, unpinned tensors only; pinned/unquantized tensors are identical in both"
+  },
+  "body_base_share": 0.7,
+  "body_bpw": 4.965,
+  "body_bpw_heuristic": 4.635,
+  "total_params": 8190735360,
+  "file_bytes_pred": 5172494336,
+  "file_bpw_pred": 5.052,
+  "histogram": {
+    "attn_k": {
+      "q6_k": 28,
+      "iq4_nl": 7,
+      "q8_0": 1
+    },
+    "attn_output": {
+      "q5_k": 15,
+      "iq4_nl": 12,
+      "q6_k": 9
+    },
+    "attn_q": {
+      "iq4_nl": 36
+    },
+    "attn_v": {
+      "q8_0": 28,
+      "q6_k": 8
+    },
+    "ffn_down": {
+      "iq4_nl": 27,
+      "q6_k": 9
+    },
+    "ffn_gate": {
+      "iq4_nl": 36
+    },
+    "ffn_up": {
+      "iq4_nl": 27,
+      "q6_k": 9
+    }
+  },
+  "prior": "layout/prior.json",
+  "pin_rule": "floor",
+  "embd_cap": "heur",
+  "generator": "auto_quant_v2.layout",
+  "generator_key": "3bf8b43e20a68f0b",
+  "quant_type": "IQ4_NL",
+  "quantize_ftype": "IQ4_NL",
+  "llama_cpp_version": "b11490",
+  "unknown_kinds": {},
+  "commit": "5d550ef5561a30307f1d436db2c6835b75395e77"
+}

layouts/AesCode-8B-IQ4_NL.tensor-types.txt

@@ -0,0 +1,254 @@
+^output\.weight$=q6_k
+^token_embd\.weight$=iq4_nl
+^blk\.0\.attn_k\.weight$=q8_0
+^blk\.0\.attn_output\.weight$=iq4_nl
+^blk\.0\.attn_q\.weight$=iq4_nl
+^blk\.0\.attn_v\.weight$=q8_0
+^blk\.0\.ffn_down\.weight$=iq4_nl
+^blk\.0\.ffn_gate\.weight$=iq4_nl
+^blk\.0\.ffn_up\.weight$=iq4_nl
+^blk\.1\.attn_k\.weight$=q6_k
+^blk\.1\.attn_output\.weight$=q6_k
+^blk\.1\.attn_q\.weight$=iq4_nl
+^blk\.1\.attn_v\.weight$=q8_0
+^blk\.1\.ffn_down\.weight$=q6_k
+^blk\.1\.ffn_gate\.weight$=iq4_nl
+^blk\.1\.ffn_up\.weight$=q6_k
+^blk\.2\.attn_k\.weight$=q6_k
+^blk\.2\.attn_output\.weight$=q6_k
+^blk\.2\.attn_q\.weight$=iq4_nl
+^blk\.2\.attn_v\.weight$=q8_0
+^blk\.2\.ffn_down\.weight$=q6_k
+^blk\.2\.ffn_gate\.weight$=iq4_nl
+^blk\.2\.ffn_up\.weight$=q6_k
+^blk\.3\.attn_k\.weight$=q6_k
+^blk\.3\.attn_output\.weight$=q6_k
+^blk\.3\.attn_q\.weight$=iq4_nl
+^blk\.3\.attn_v\.weight$=q8_0
+^blk\.3\.ffn_down\.weight$=q6_k
+^blk\.3\.ffn_gate\.weight$=iq4_nl
+^blk\.3\.ffn_up\.weight$=q6_k
+^blk\.4\.attn_k\.weight$=q6_k
+^blk\.4\.attn_output\.weight$=q6_k
+^blk\.4\.attn_q\.weight$=iq4_nl
+^blk\.4\.attn_v\.weight$=q8_0
+^blk\.4\.ffn_down\.weight$=q6_k
+^blk\.4\.ffn_gate\.weight$=iq4_nl
+^blk\.4\.ffn_up\.weight$=q6_k
+^blk\.5\.attn_k\.weight$=q6_k
+^blk\.5\.attn_output\.weight$=q6_k
+^blk\.5\.attn_q\.weight$=iq4_nl
+^blk\.5\.attn_v\.weight$=q8_0
+^blk\.5\.ffn_down\.weight$=q6_k
+^blk\.5\.ffn_gate\.weight$=iq4_nl
+^blk\.5\.ffn_up\.weight$=q6_k
+^blk\.6\.attn_k\.weight$=q6_k
+^blk\.6\.attn_output\.weight$=q6_k
+^blk\.6\.attn_q\.weight$=iq4_nl
+^blk\.6\.attn_v\.weight$=q8_0
+^blk\.6\.ffn_down\.weight$=q6_k
+^blk\.6\.ffn_gate\.weight$=iq4_nl
+^blk\.6\.ffn_up\.weight$=q6_k
+^blk\.7\.attn_k\.weight$=q6_k
+^blk\.7\.attn_output\.weight$=q6_k
+^blk\.7\.attn_q\.weight$=iq4_nl
+^blk\.7\.attn_v\.weight$=q8_0
+^blk\.7\.ffn_down\.weight$=q6_k
+^blk\.7\.ffn_gate\.weight$=iq4_nl
+^blk\.7\.ffn_up\.weight$=q6_k
+^blk\.8\.attn_k\.weight$=q6_k
+^blk\.8\.attn_output\.weight$=q6_k
+^blk\.8\.attn_q\.weight$=iq4_nl
+^blk\.8\.attn_v\.weight$=q8_0
+^blk\.8\.ffn_down\.weight$=q6_k
+^blk\.8\.ffn_gate\.weight$=iq4_nl
+^blk\.8\.ffn_up\.weight$=q6_k
+^blk\.9\.attn_k\.weight$=q6_k
+^blk\.9\.attn_output\.weight$=q6_k
+^blk\.9\.attn_q\.weight$=iq4_nl
+^blk\.9\.attn_v\.weight$=q8_0
+^blk\.9\.ffn_down\.weight$=iq4_nl
+^blk\.9\.ffn_gate\.weight$=iq4_nl
+^blk\.9\.ffn_up\.weight$=iq4_nl
+^blk\.10\.attn_k\.weight$=q6_k
+^blk\.10\.attn_output\.weight$=q5_k
+^blk\.10\.attn_q\.weight$=iq4_nl
+^blk\.10\.attn_v\.weight$=q8_0
+^blk\.10\.ffn_down\.weight$=iq4_nl
+^blk\.10\.ffn_gate\.weight$=iq4_nl
+^blk\.10\.ffn_up\.weight$=iq4_nl
+^blk\.11\.attn_k\.weight$=q6_k
+^blk\.11\.attn_output\.weight$=q5_k
+^blk\.11\.attn_q\.weight$=iq4_nl
+^blk\.11\.attn_v\.weight$=q8_0
+^blk\.11\.ffn_down\.weight$=iq4_nl
+^blk\.11\.ffn_gate\.weight$=iq4_nl
+^blk\.11\.ffn_up\.weight$=iq4_nl
+^blk\.12\.attn_k\.weight$=q6_k
+^blk\.12\.attn_output\.weight$=q5_k
+^blk\.12\.attn_q\.weight$=iq4_nl
+^blk\.12\.attn_v\.weight$=q8_0
+^blk\.12\.ffn_down\.weight$=iq4_nl
+^blk\.12\.ffn_gate\.weight$=iq4_nl
+^blk\.12\.ffn_up\.weight$=iq4_nl
+^blk\.13\.attn_k\.weight$=q6_k
+^blk\.13\.attn_output\.weight$=q5_k
+^blk\.13\.attn_q\.weight$=iq4_nl
+^blk\.13\.attn_v\.weight$=q8_0
+^blk\.13\.ffn_down\.weight$=iq4_nl
+^blk\.13\.ffn_gate\.weight$=iq4_nl
+^blk\.13\.ffn_up\.weight$=iq4_nl
+^blk\.14\.attn_k\.weight$=q6_k
+^blk\.14\.attn_output\.weight$=q5_k
+^blk\.14\.attn_q\.weight$=iq4_nl
+^blk\.14\.attn_v\.weight$=q8_0
+^blk\.14\.ffn_down\.weight$=iq4_nl
+^blk\.14\.ffn_gate\.weight$=iq4_nl
+^blk\.14\.ffn_up\.weight$=iq4_nl
+^blk\.15\.attn_k\.weight$=q6_k
+^blk\.15\.attn_output\.weight$=q5_k
+^blk\.15\.attn_q\.weight$=iq4_nl
+^blk\.15\.attn_v\.weight$=q8_0
+^blk\.15\.ffn_down\.weight$=iq4_nl
+^blk\.15\.ffn_gate\.weight$=iq4_nl
+^blk\.15\.ffn_up\.weight$=iq4_nl
+^blk\.16\.attn_k\.weight$=q6_k
+^blk\.16\.attn_output\.weight$=q5_k
+^blk\.16\.attn_q\.weight$=iq4_nl
+^blk\.16\.attn_v\.weight$=q8_0
+^blk\.16\.ffn_down\.weight$=iq4_nl
+^blk\.16\.ffn_gate\.weight$=iq4_nl
+^blk\.16\.ffn_up\.weight$=iq4_nl
+^blk\.17\.attn_k\.weight$=q6_k
+^blk\.17\.attn_output\.weight$=q5_k
+^blk\.17\.attn_q\.weight$=iq4_nl
+^blk\.17\.attn_v\.weight$=q8_0
+^blk\.17\.ffn_down\.weight$=iq4_nl
+^blk\.17\.ffn_gate\.weight$=iq4_nl
+^blk\.17\.ffn_up\.weight$=iq4_nl
+^blk\.18\.attn_k\.weight$=q6_k
+^blk\.18\.attn_output\.weight$=q5_k
+^blk\.18\.attn_q\.weight$=iq4_nl
+^blk\.18\.attn_v\.weight$=q8_0
+^blk\.18\.ffn_down\.weight$=iq4_nl
+^blk\.18\.ffn_gate\.weight$=iq4_nl
+^blk\.18\.ffn_up\.weight$=iq4_nl
+^blk\.19\.attn_k\.weight$=q6_k
+^blk\.19\.attn_output\.weight$=q5_k
+^blk\.19\.attn_q\.weight$=iq4_nl
+^blk\.19\.attn_v\.weight$=q8_0
+^blk\.19\.ffn_down\.weight$=iq4_nl
+^blk\.19\.ffn_gate\.weight$=iq4_nl
+^blk\.19\.ffn_up\.weight$=iq4_nl
+^blk\.20\.attn_k\.weight$=q6_k
+^blk\.20\.attn_output\.weight$=q5_k
+^blk\.20\.attn_q\.weight$=iq4_nl
+^blk\.20\.attn_v\.weight$=q8_0
+^blk\.20\.ffn_down\.weight$=iq4_nl
+^blk\.20\.ffn_gate\.weight$=iq4_nl
+^blk\.20\.ffn_up\.weight$=iq4_nl
+^blk\.21\.attn_k\.weight$=q6_k
+^blk\.21\.attn_output\.weight$=q5_k
+^blk\.21\.attn_q\.weight$=iq4_nl
+^blk\.21\.attn_v\.weight$=q8_0
+^blk\.21\.ffn_down\.weight$=iq4_nl
+^blk\.21\.ffn_gate\.weight$=iq4_nl
+^blk\.21\.ffn_up\.weight$=iq4_nl
+^blk\.22\.attn_k\.weight$=q6_k
+^blk\.22\.attn_output\.weight$=q5_k
+^blk\.22\.attn_q\.weight$=iq4_nl
+^blk\.22\.attn_v\.weight$=q8_0
+^blk\.22\.ffn_down\.weight$=iq4_nl
+^blk\.22\.ffn_gate\.weight$=iq4_nl
+^blk\.22\.ffn_up\.weight$=iq4_nl
+^blk\.23\.attn_k\.weight$=q6_k
+^blk\.23\.attn_output\.weight$=q5_k
+^blk\.23\.attn_q\.weight$=iq4_nl
+^blk\.23\.attn_v\.weight$=q8_0
+^blk\.23\.ffn_down\.weight$=iq4_nl
+^blk\.23\.ffn_gate\.weight$=iq4_nl
+^blk\.23\.ffn_up\.weight$=iq4_nl
+^blk\.24\.attn_k\.weight$=q6_k
+^blk\.24\.attn_output\.weight$=q5_k
+^blk\.24\.attn_q\.weight$=iq4_nl
+^blk\.24\.attn_v\.weight$=q8_0
+^blk\.24\.ffn_down\.weight$=iq4_nl
+^blk\.24\.ffn_gate\.weight$=iq4_nl
+^blk\.24\.ffn_up\.weight$=iq4_nl
+^blk\.25\.attn_k\.weight$=q6_k
+^blk\.25\.attn_output\.weight$=iq4_nl
+^blk\.25\.attn_q\.weight$=iq4_nl
+^blk\.25\.attn_v\.weight$=q8_0
+^blk\.25\.ffn_down\.weight$=iq4_nl
+^blk\.25\.ffn_gate\.weight$=iq4_nl
+^blk\.25\.ffn_up\.weight$=iq4_nl
+^blk\.26\.attn_k\.weight$=q6_k
+^blk\.26\.attn_output\.weight$=iq4_nl
+^blk\.26\.attn_q\.weight$=iq4_nl
+^blk\.26\.attn_v\.weight$=q8_0
+^blk\.26\.ffn_down\.weight$=iq4_nl
+^blk\.26\.ffn_gate\.weight$=iq4_nl
+^blk\.26\.ffn_up\.weight$=iq4_nl
+^blk\.27\.attn_k\.weight$=q6_k
+^blk\.27\.attn_output\.weight$=iq4_nl
+^blk\.27\.attn_q\.weight$=iq4_nl
+^blk\.27\.attn_v\.weight$=q6_k
+^blk\.27\.ffn_down\.weight$=iq4_nl
+^blk\.27\.ffn_gate\.weight$=iq4_nl
+^blk\.27\.ffn_up\.weight$=iq4_nl
+^blk\.28\.attn_k\.weight$=q6_k
+^blk\.28\.attn_output\.weight$=iq4_nl
+^blk\.28\.attn_q\.weight$=iq4_nl
+^blk\.28\.attn_v\.weight$=q6_k
+^blk\.28\.ffn_down\.weight$=iq4_nl
+^blk\.28\.ffn_gate\.weight$=iq4_nl
+^blk\.28\.ffn_up\.weight$=iq4_nl
+^blk\.29\.attn_k\.weight$=iq4_nl
+^blk\.29\.attn_output\.weight$=iq4_nl
+^blk\.29\.attn_q\.weight$=iq4_nl
+^blk\.29\.attn_v\.weight$=q6_k
+^blk\.29\.ffn_down\.weight$=iq4_nl
+^blk\.29\.ffn_gate\.weight$=iq4_nl
+^blk\.29\.ffn_up\.weight$=iq4_nl
+^blk\.30\.attn_k\.weight$=iq4_nl
+^blk\.30\.attn_output\.weight$=iq4_nl
+^blk\.30\.attn_q\.weight$=iq4_nl
+^blk\.30\.attn_v\.weight$=q6_k
+^blk\.30\.ffn_down\.weight$=iq4_nl
+^blk\.30\.ffn_gate\.weight$=iq4_nl
+^blk\.30\.ffn_up\.weight$=iq4_nl
+^blk\.31\.attn_k\.weight$=iq4_nl
+^blk\.31\.attn_output\.weight$=iq4_nl
+^blk\.31\.attn_q\.weight$=iq4_nl
+^blk\.31\.attn_v\.weight$=q6_k
+^blk\.31\.ffn_down\.weight$=iq4_nl
+^blk\.31\.ffn_gate\.weight$=iq4_nl
+^blk\.31\.ffn_up\.weight$=iq4_nl
+^blk\.32\.attn_k\.weight$=iq4_nl
+^blk\.32\.attn_output\.weight$=iq4_nl
+^blk\.32\.attn_q\.weight$=iq4_nl
+^blk\.32\.attn_v\.weight$=q6_k
+^blk\.32\.ffn_down\.weight$=iq4_nl
+^blk\.32\.ffn_gate\.weight$=iq4_nl
+^blk\.32\.ffn_up\.weight$=iq4_nl
+^blk\.33\.attn_k\.weight$=iq4_nl
+^blk\.33\.attn_output\.weight$=iq4_nl
+^blk\.33\.attn_q\.weight$=iq4_nl
+^blk\.33\.attn_v\.weight$=q6_k
+^blk\.33\.ffn_down\.weight$=iq4_nl
+^blk\.33\.ffn_gate\.weight$=iq4_nl
+^blk\.33\.ffn_up\.weight$=iq4_nl
+^blk\.34\.attn_k\.weight$=iq4_nl
+^blk\.34\.attn_output\.weight$=iq4_nl
+^blk\.34\.attn_q\.weight$=iq4_nl
+^blk\.34\.attn_v\.weight$=q6_k
+^blk\.34\.ffn_down\.weight$=iq4_nl
+^blk\.34\.ffn_gate\.weight$=iq4_nl
+^blk\.34\.ffn_up\.weight$=iq4_nl
+^blk\.35\.attn_k\.weight$=iq4_nl
+^blk\.35\.attn_output\.weight$=iq4_nl
+^blk\.35\.attn_q\.weight$=iq4_nl
+^blk\.35\.attn_v\.weight$=q8_0
+^blk\.35\.ffn_down\.weight$=q6_k
+^blk\.35\.ffn_gate\.weight$=iq4_nl
+^blk\.35\.ffn_up\.weight$=q6_k

layouts/AesCode-8B-IQ4_XS.layout.json

@@ -0,0 +1,86 @@
+{
+  "model": "/media/m2teamgroup/models/AesCode-8B-GGUF/AesCode-8B-bf16.gguf",
+  "imatrix": "/media/m2teamgroup/models/AesCode-8B-GGUF/AesCode-8B-imatrix.gguf",
+  "ftype": "IQ4_XS",
+  "base_type": "iq4_xs",
+  "family": "iq",
+  "recipe": "ident",
+  "variant": "shr0.90",
+  "auto_crush": false,
+  "rung_mode": "share",
+  "tier": "S",
+  "rung_share": 0.9,
+  "crush_pass_bytes": null,
+  "embd_shares": {
+    "output.weight": {
+      "share": 0.1095,
+      "type": "q6_k"
+    },
+    "token_embd.weight": {
+      "share": 0.0709,
+      "type": "iq4_xs"
+    }
+  },
+  "arch": "dense",
+  "tied": false,
+  "n_layer": 36,
+  "body_rule": true,
+  "pins": {},
+  "shape_pins": {},
+  "fixed_pins": {},
+  "embeddings": {
+    "output.weight": "q6_k",
+    "token_embd.weight": "iq4_xs"
+  },
+  "bytes": {
+    "heuristic_body+embd": 4661997568,
+    "map_body+embd": 4680839168,
+    "ratio": 1.004,
+    "note": "quantizable, unpinned tensors only; pinned/unquantized tensors are identical in both"
+  },
+  "body_base_share": 0.9,
+  "body_bpw": 4.423,
+  "body_bpw_heuristic": 4.401,
+  "total_params": 8190735360,
+  "file_bytes_pred": 4682072064,
+  "file_bpw_pred": 4.573,
+  "histogram": {
+    "attn_k": {
+      "q6_k": 19,
+      "iq4_xs": 11,
+      "q4_k": 5,
+      "q5_k": 1
+    },
+    "attn_output": {
+      "iq4_xs": 28,
+      "q6_k": 8
+    },
+    "attn_q": {
+      "iq4_xs": 36
+    },
+    "attn_v": {
+      "q8_0": 28,
+      "q6_k": 8
+    },
+    "ffn_down": {
+      "iq4_xs": 36
+    },
+    "ffn_gate": {
+      "iq4_xs": 36
+    },
+    "ffn_up": {
+      "iq4_xs": 35,
+      "q6_k": 1
+    }
+  },
+  "prior": "layout/prior.json",
+  "pin_rule": "floor",
+  "embd_cap": "heur",
+  "generator": "auto_quant_v2.layout",
+  "generator_key": "3bf8b43e20a68f0b",
+  "quant_type": "IQ4_XS",
+  "quantize_ftype": "IQ4_XS",
+  "llama_cpp_version": "b11490",
+  "unknown_kinds": {},
+  "commit": "5d550ef5561a30307f1d436db2c6835b75395e77"
+}

layouts/AesCode-8B-IQ4_XS.tensor-types.txt

@@ -0,0 +1,254 @@
+^output\.weight$=q6_k
+^token_embd\.weight$=iq4_xs
+^blk\.0\.attn_k\.weight$=q6_k
+^blk\.0\.attn_output\.weight$=iq4_xs
+^blk\.0\.attn_q\.weight$=iq4_xs
+^blk\.0\.attn_v\.weight$=q8_0
+^blk\.0\.ffn_down\.weight$=iq4_xs
+^blk\.0\.ffn_gate\.weight$=iq4_xs
+^blk\.0\.ffn_up\.weight$=iq4_xs
+^blk\.1\.attn_k\.weight$=q5_k
+^blk\.1\.attn_output\.weight$=q6_k
+^blk\.1\.attn_q\.weight$=iq4_xs
+^blk\.1\.attn_v\.weight$=q8_0
+^blk\.1\.ffn_down\.weight$=iq4_xs
+^blk\.1\.ffn_gate\.weight$=iq4_xs
+^blk\.1\.ffn_up\.weight$=iq4_xs
+^blk\.2\.attn_k\.weight$=q4_k
+^blk\.2\.attn_output\.weight$=q6_k
+^blk\.2\.attn_q\.weight$=iq4_xs
+^blk\.2\.attn_v\.weight$=q8_0
+^blk\.2\.ffn_down\.weight$=iq4_xs
+^blk\.2\.ffn_gate\.weight$=iq4_xs
+^blk\.2\.ffn_up\.weight$=iq4_xs
+^blk\.3\.attn_k\.weight$=q4_k
+^blk\.3\.attn_output\.weight$=q6_k
+^blk\.3\.attn_q\.weight$=iq4_xs
+^blk\.3\.attn_v\.weight$=q8_0
+^blk\.3\.ffn_down\.weight$=iq4_xs
+^blk\.3\.ffn_gate\.weight$=iq4_xs
+^blk\.3\.ffn_up\.weight$=iq4_xs
+^blk\.4\.attn_k\.weight$=q4_k
+^blk\.4\.attn_output\.weight$=q6_k
+^blk\.4\.attn_q\.weight$=iq4_xs
+^blk\.4\.attn_v\.weight$=q8_0
+^blk\.4\.ffn_down\.weight$=iq4_xs
+^blk\.4\.ffn_gate\.weight$=iq4_xs
+^blk\.4\.ffn_up\.weight$=iq4_xs
+^blk\.5\.attn_k\.weight$=q4_k
+^blk\.5\.attn_output\.weight$=q6_k
+^blk\.5\.attn_q\.weight$=iq4_xs
+^blk\.5\.attn_v\.weight$=q8_0
+^blk\.5\.ffn_down\.weight$=iq4_xs
+^blk\.5\.ffn_gate\.weight$=iq4_xs
+^blk\.5\.ffn_up\.weight$=iq4_xs
+^blk\.6\.attn_k\.weight$=q4_k
+^blk\.6\.attn_output\.weight$=q6_k
+^blk\.6\.attn_q\.weight$=iq4_xs
+^blk\.6\.attn_v\.weight$=q8_0
+^blk\.6\.ffn_down\.weight$=iq4_xs
+^blk\.6\.ffn_gate\.weight$=iq4_xs
+^blk\.6\.ffn_up\.weight$=iq4_xs
+^blk\.7\.attn_k\.weight$=iq4_xs
+^blk\.7\.attn_output\.weight$=q6_k
+^blk\.7\.attn_q\.weight$=iq4_xs
+^blk\.7\.attn_v\.weight$=q8_0
+^blk\.7\.ffn_down\.weight$=iq4_xs
+^blk\.7\.ffn_gate\.weight$=iq4_xs
+^blk\.7\.ffn_up\.weight$=iq4_xs
+^blk\.8\.attn_k\.weight$=iq4_xs
+^blk\.8\.attn_output\.weight$=q6_k
+^blk\.8\.attn_q\.weight$=iq4_xs
+^blk\.8\.attn_v\.weight$=q8_0
+^blk\.8\.ffn_down\.weight$=iq4_xs
+^blk\.8\.ffn_gate\.weight$=iq4_xs
+^blk\.8\.ffn_up\.weight$=iq4_xs
+^blk\.9\.attn_k\.weight$=q6_k
+^blk\.9\.attn_output\.weight$=iq4_xs
+^blk\.9\.attn_q\.weight$=iq4_xs
+^blk\.9\.attn_v\.weight$=q8_0
+^blk\.9\.ffn_down\.weight$=iq4_xs
+^blk\.9\.ffn_gate\.weight$=iq4_xs
+^blk\.9\.ffn_up\.weight$=iq4_xs
+^blk\.10\.attn_k\.weight$=q6_k
+^blk\.10\.attn_output\.weight$=iq4_xs
+^blk\.10\.attn_q\.weight$=iq4_xs
+^blk\.10\.attn_v\.weight$=q8_0
+^blk\.10\.ffn_down\.weight$=iq4_xs
+^blk\.10\.ffn_gate\.weight$=iq4_xs
+^blk\.10\.ffn_up\.weight$=iq4_xs
+^blk\.11\.attn_k\.weight$=q6_k
+^blk\.11\.attn_output\.weight$=iq4_xs
+^blk\.11\.attn_q\.weight$=iq4_xs
+^blk\.11\.attn_v\.weight$=q8_0
+^blk\.11\.ffn_down\.weight$=iq4_xs
+^blk\.11\.ffn_gate\.weight$=iq4_xs
+^blk\.11\.ffn_up\.weight$=iq4_xs
+^blk\.12\.attn_k\.weight$=q6_k
+^blk\.12\.attn_output\.weight$=iq4_xs
+^blk\.12\.attn_q\.weight$=iq4_xs
+^blk\.12\.attn_v\.weight$=q8_0
+^blk\.12\.ffn_down\.weight$=iq4_xs
+^blk\.12\.ffn_gate\.weight$=iq4_xs
+^blk\.12\.ffn_up\.weight$=iq4_xs
+^blk\.13\.attn_k\.weight$=q6_k
+^blk\.13\.attn_output\.weight$=iq4_xs
+^blk\.13\.attn_q\.weight$=iq4_xs
+^blk\.13\.attn_v\.weight$=q8_0
+^blk\.13\.ffn_down\.weight$=iq4_xs
+^blk\.13\.ffn_gate\.weight$=iq4_xs
+^blk\.13\.ffn_up\.weight$=iq4_xs
+^blk\.14\.attn_k\.weight$=q6_k
+^blk\.14\.attn_output\.weight$=iq4_xs
+^blk\.14\.attn_q\.weight$=iq4_xs
+^blk\.14\.attn_v\.weight$=q8_0
+^blk\.14\.ffn_down\.weight$=iq4_xs
+^blk\.14\.ffn_gate\.weight$=iq4_xs
+^blk\.14\.ffn_up\.weight$=iq4_xs
+^blk\.15\.attn_k\.weight$=q6_k
+^blk\.15\.attn_output\.weight$=iq4_xs
+^blk\.15\.attn_q\.weight$=iq4_xs
+^blk\.15\.attn_v\.weight$=q8_0
+^blk\.15\.ffn_down\.weight$=iq4_xs
+^blk\.15\.ffn_gate\.weight$=iq4_xs
+^blk\.15\.ffn_up\.weight$=iq4_xs
+^blk\.16\.attn_k\.weight$=q6_k
+^blk\.16\.attn_output\.weight$=iq4_xs
+^blk\.16\.attn_q\.weight$=iq4_xs
+^blk\.16\.attn_v\.weight$=q8_0
+^blk\.16\.ffn_down\.weight$=iq4_xs
+^blk\.16\.ffn_gate\.weight$=iq4_xs
+^blk\.16\.ffn_up\.weight$=iq4_xs
+^blk\.17\.attn_k\.weight$=q6_k
+^blk\.17\.attn_output\.weight$=iq4_xs
+^blk\.17\.attn_q\.weight$=iq4_xs
+^blk\.17\.attn_v\.weight$=q8_0
+^blk\.17\.ffn_down\.weight$=iq4_xs
+^blk\.17\.ffn_gate\.weight$=iq4_xs
+^blk\.17\.ffn_up\.weight$=iq4_xs
+^blk\.18\.attn_k\.weight$=q6_k
+^blk\.18\.attn_output\.weight$=iq4_xs
+^blk\.18\.attn_q\.weight$=iq4_xs
+^blk\.18\.attn_v\.weight$=q8_0
+^blk\.18\.ffn_down\.weight$=iq4_xs
+^blk\.18\.ffn_gate\.weight$=iq4_xs
+^blk\.18\.ffn_up\.weight$=iq4_xs
+^blk\.19\.attn_k\.weight$=q6_k
+^blk\.19\.attn_output\.weight$=iq4_xs
+^blk\.19\.attn_q\.weight$=iq4_xs
+^blk\.19\.attn_v\.weight$=q8_0
+^blk\.19\.ffn_down\.weight$=iq4_xs
+^blk\.19\.ffn_gate\.weight$=iq4_xs
+^blk\.19\.ffn_up\.weight$=iq4_xs
+^blk\.20\.attn_k\.weight$=q6_k
+^blk\.20\.attn_output\.weight$=iq4_xs
+^blk\.20\.attn_q\.weight$=iq4_xs
+^blk\.20\.attn_v\.weight$=q8_0
+^blk\.20\.ffn_down\.weight$=iq4_xs
+^blk\.20\.ffn_gate\.weight$=iq4_xs
+^blk\.20\.ffn_up\.weight$=iq4_xs
+^blk\.21\.attn_k\.weight$=q6_k
+^blk\.21\.attn_output\.weight$=iq4_xs
+^blk\.21\.attn_q\.weight$=iq4_xs
+^blk\.21\.attn_v\.weight$=q8_0
+^blk\.21\.ffn_down\.weight$=iq4_xs
+^blk\.21\.ffn_gate\.weight$=iq4_xs
+^blk\.21\.ffn_up\.weight$=iq4_xs
+^blk\.22\.attn_k\.weight$=q6_k
+^blk\.22\.attn_output\.weight$=iq4_xs
+^blk\.22\.attn_q\.weight$=iq4_xs
+^blk\.22\.attn_v\.weight$=q8_0
+^blk\.22\.ffn_down\.weight$=iq4_xs
+^blk\.22\.ffn_gate\.weight$=iq4_xs
+^blk\.22\.ffn_up\.weight$=iq4_xs
+^blk\.23\.attn_k\.weight$=q6_k
+^blk\.23\.attn_output\.weight$=iq4_xs
+^blk\.23\.attn_q\.weight$=iq4_xs
+^blk\.23\.attn_v\.weight$=q8_0
+^blk\.23\.ffn_down\.weight$=iq4_xs
+^blk\.23\.ffn_gate\.weight$=iq4_xs
+^blk\.23\.ffn_up\.weight$=iq4_xs
+^blk\.24\.attn_k\.weight$=q6_k
+^blk\.24\.attn_output\.weight$=iq4_xs
+^blk\.24\.attn_q\.weight$=iq4_xs
+^blk\.24\.attn_v\.weight$=q8_0
+^blk\.24\.ffn_down\.weight$=iq4_xs
+^blk\.24\.ffn_gate\.weight$=iq4_xs
+^blk\.24\.ffn_up\.weight$=iq4_xs
+^blk\.25\.attn_k\.weight$=q6_k
+^blk\.25\.attn_output\.weight$=iq4_xs
+^blk\.25\.attn_q\.weight$=iq4_xs
+^blk\.25\.attn_v\.weight$=q8_0
+^blk\.25\.ffn_down\.weight$=iq4_xs
+^blk\.25\.ffn_gate\.weight$=iq4_xs
+^blk\.25\.ffn_up\.weight$=iq4_xs
+^blk\.26\.attn_k\.weight$=q6_k
+^blk\.26\.attn_output\.weight$=iq4_xs
+^blk\.26\.attn_q\.weight$=iq4_xs
+^blk\.26\.attn_v\.weight$=q8_0
+^blk\.26\.ffn_down\.weight$=iq4_xs
+^blk\.26\.ffn_gate\.weight$=iq4_xs
+^blk\.26\.ffn_up\.weight$=iq4_xs
+^blk\.27\.attn_k\.weight$=iq4_xs
+^blk\.27\.attn_output\.weight$=iq4_xs
+^blk\.27\.attn_q\.weight$=iq4_xs
+^blk\.27\.attn_v\.weight$=q6_k
+^blk\.27\.ffn_down\.weight$=iq4_xs
+^blk\.27\.ffn_gate\.weight$=iq4_xs
+^blk\.27\.ffn_up\.weight$=iq4_xs
+^blk\.28\.attn_k\.weight$=iq4_xs
+^blk\.28\.attn_output\.weight$=iq4_xs
+^blk\.28\.attn_q\.weight$=iq4_xs
+^blk\.28\.attn_v\.weight$=q6_k
+^blk\.28\.ffn_down\.weight$=iq4_xs
+^blk\.28\.ffn_gate\.weight$=iq4_xs
+^blk\.28\.ffn_up\.weight$=iq4_xs
+^blk\.29\.attn_k\.weight$=iq4_xs
+^blk\.29\.attn_output\.weight$=iq4_xs
+^blk\.29\.attn_q\.weight$=iq4_xs
+^blk\.29\.attn_v\.weight$=q6_k
+^blk\.29\.ffn_down\.weight$=iq4_xs
+^blk\.29\.ffn_gate\.weight$=iq4_xs
+^blk\.29\.ffn_up\.weight$=iq4_xs
+^blk\.30\.attn_k\.weight$=iq4_xs
+^blk\.30\.attn_output\.weight$=iq4_xs
+^blk\.30\.attn_q\.weight$=iq4_xs
+^blk\.30\.attn_v\.weight$=q6_k
+^blk\.30\.ffn_down\.weight$=iq4_xs
+^blk\.30\.ffn_gate\.weight$=iq4_xs
+^blk\.30\.ffn_up\.weight$=iq4_xs
+^blk\.31\.attn_k\.weight$=iq4_xs
+^blk\.31\.attn_output\.weight$=iq4_xs
+^blk\.31\.attn_q\.weight$=iq4_xs
+^blk\.31\.attn_v\.weight$=q6_k
+^blk\.31\.ffn_down\.weight$=iq4_xs
+^blk\.31\.ffn_gate\.weight$=iq4_xs
+^blk\.31\.ffn_up\.weight$=iq4_xs
+^blk\.32\.attn_k\.weight$=iq4_xs
+^blk\.32\.attn_output\.weight$=iq4_xs
+^blk\.32\.attn_q\.weight$=iq4_xs
+^blk\.32\.attn_v\.weight$=q6_k
+^blk\.32\.ffn_down\.weight$=iq4_xs
+^blk\.32\.ffn_gate\.weight$=iq4_xs
+^blk\.32\.ffn_up\.weight$=iq4_xs
+^blk\.33\.attn_k\.weight$=iq4_xs
+^blk\.33\.attn_output\.weight$=iq4_xs
+^blk\.33\.attn_q\.weight$=iq4_xs
+^blk\.33\.attn_v\.weight$=q6_k
+^blk\.33\.ffn_down\.weight$=iq4_xs
+^blk\.33\.ffn_gate\.weight$=iq4_xs
+^blk\.33\.ffn_up\.weight$=iq4_xs
+^blk\.34\.attn_k\.weight$=iq4_xs
+^blk\.34\.attn_output\.weight$=iq4_xs
+^blk\.34\.attn_q\.weight$=iq4_xs
+^blk\.34\.attn_v\.weight$=q6_k
+^blk\.34\.ffn_down\.weight$=iq4_xs
+^blk\.34\.ffn_gate\.weight$=iq4_xs
+^blk\.34\.ffn_up\.weight$=iq4_xs
+^blk\.35\.attn_k\.weight$=iq4_xs
+^blk\.35\.attn_output\.weight$=iq4_xs
+^blk\.35\.attn_q\.weight$=iq4_xs
+^blk\.35\.attn_v\.weight$=q8_0
+^blk\.35\.ffn_down\.weight$=iq4_xs
+^blk\.35\.ffn_gate\.weight$=iq4_xs
+^blk\.35\.ffn_up\.weight$=q6_k

layouts/AesCode-8B-Q2_K.layout.json

@@ -0,0 +1,86 @@
+{
+  "model": "/media/m2teamgroup/models/AesCode-8B-GGUF/AesCode-8B-bf16.gguf",
+  "imatrix": "/media/m2teamgroup/models/AesCode-8B-GGUF/AesCode-8B-imatrix.gguf",
+  "ftype": "Q2_K",
+  "base_type": "q2_k",
+  "family": "k",
+  "recipe": "ident",
+  "variant": "shr0.70",
+  "auto_crush": false,
+  "rung_mode": "share",
+  "tier": "M",
+  "rung_share": 0.7,
+  "crush_pass_bytes": null,
+  "embd_shares": {
+    "output.weight": {
+      "share": 0.1491,
+      "type": "q5_k"
+    },
+    "token_embd.weight": {
+      "share": 0.0596,
+      "type": "q3_k"
+    }
+  },
+  "arch": "dense",
+  "tied": false,
+  "n_layer": 36,
+  "body_rule": true,
+  "pins": {},
+  "shape_pins": {},
+  "fixed_pins": {},
+  "embeddings": {
+    "output.weight": "q5_k",
+    "token_embd.weight": "q3_k"
+  },
+  "bytes": {
+    "heuristic_body+embd": 3423178752,
+    "map_body+embd": 3292385280,
+    "ratio": 0.9618,
+    "note": "quantizable, unpinned tensors only; pinned/unquantized tensors are identical in both"
+  },
+  "body_base_share": 0.7,
+  "body_bpw": 2.991,
+  "body_bpw_heuristic": 3.12,
+  "total_params": 8190735360,
+  "file_bytes_pred": 3293618176,
+  "file_bpw_pred": 3.217,
+  "histogram": {
+    "attn_k": {
+      "q4_k": 27,
+      "q2_k": 9
+    },
+    "attn_output": {
+      "q2_k": 22,
+      "q4_k": 8,
+      "q3_k": 6
+    },
+    "attn_q": {
+      "q2_k": 36
+    },
+    "attn_v": {
+      "q4_k": 32,
+      "q5_k": 4
+    },
+    "ffn_down": {
+      "q2_k": 27,
+      "q4_k": 9
+    },
+    "ffn_gate": {
+      "q2_k": 36
+    },
+    "ffn_up": {
+      "q2_k": 27,
+      "q4_k": 9
+    }
+  },
+  "prior": "layout/prior.json",
+  "pin_rule": "floor",
+  "embd_cap": "heur",
+  "generator": "auto_quant_v2.layout",
+  "generator_key": "3bf8b43e20a68f0b",
+  "quant_type": "Q2_K",
+  "quantize_ftype": "Q2_K",
+  "llama_cpp_version": "b11490",
+  "unknown_kinds": {},
+  "commit": "5d550ef5561a30307f1d436db2c6835b75395e77"
+}

layouts/AesCode-8B-Q2_K.tensor-types.txt

@@ -0,0 +1,254 @@
+^output\.weight$=q5_k
+^token_embd\.weight$=q3_k
+^blk\.0\.attn_k\.weight$=q4_k
+^blk\.0\.attn_output\.weight$=q2_k
+^blk\.0\.attn_q\.weight$=q2_k
+^blk\.0\.attn_v\.weight$=q4_k
+^blk\.0\.ffn_down\.weight$=q2_k
+^blk\.0\.ffn_gate\.weight$=q2_k
+^blk\.0\.ffn_up\.weight$=q2_k
+^blk\.1\.attn_k\.weight$=q4_k
+^blk\.1\.attn_output\.weight$=q4_k
+^blk\.1\.attn_q\.weight$=q2_k
+^blk\.1\.attn_v\.weight$=q5_k
+^blk\.1\.ffn_down\.weight$=q4_k
+^blk\.1\.ffn_gate\.weight$=q2_k
+^blk\.1\.ffn_up\.weight$=q4_k
+^blk\.2\.attn_k\.weight$=q4_k
+^blk\.2\.attn_output\.weight$=q4_k
+^blk\.2\.attn_q\.weight$=q2_k
+^blk\.2\.attn_v\.weight$=q5_k
+^blk\.2\.ffn_down\.weight$=q4_k
+^blk\.2\.ffn_gate\.weight$=q2_k
+^blk\.2\.ffn_up\.weight$=q4_k
+^blk\.3\.attn_k\.weight$=q4_k
+^blk\.3\.attn_output\.weight$=q4_k
+^blk\.3\.attn_q\.weight$=q2_k
+^blk\.3\.attn_v\.weight$=q5_k
+^blk\.3\.ffn_down\.weight$=q4_k
+^blk\.3\.ffn_gate\.weight$=q2_k
+^blk\.3\.ffn_up\.weight$=q4_k
+^blk\.4\.attn_k\.weight$=q4_k
+^blk\.4\.attn_output\.weight$=q4_k
+^blk\.4\.attn_q\.weight$=q2_k
+^blk\.4\.attn_v\.weight$=q5_k
+^blk\.4\.ffn_down\.weight$=q4_k
+^blk\.4\.ffn_gate\.weight$=q2_k
+^blk\.4\.ffn_up\.weight$=q4_k
+^blk\.5\.attn_k\.weight$=q4_k
+^blk\.5\.attn_output\.weight$=q4_k
+^blk\.5\.attn_q\.weight$=q2_k
+^blk\.5\.attn_v\.weight$=q4_k
+^blk\.5\.ffn_down\.weight$=q4_k
+^blk\.5\.ffn_gate\.weight$=q2_k
+^blk\.5\.ffn_up\.weight$=q4_k
+^blk\.6\.attn_k\.weight$=q4_k
+^blk\.6\.attn_output\.weight$=q4_k
+^blk\.6\.attn_q\.weight$=q2_k
+^blk\.6\.attn_v\.weight$=q4_k
+^blk\.6\.ffn_down\.weight$=q4_k
+^blk\.6\.ffn_gate\.weight$=q2_k
+^blk\.6\.ffn_up\.weight$=q4_k
+^blk\.7\.attn_k\.weight$=q4_k
+^blk\.7\.attn_output\.weight$=q4_k
+^blk\.7\.attn_q\.weight$=q2_k
+^blk\.7\.attn_v\.weight$=q4_k
+^blk\.7\.ffn_down\.weight$=q4_k
+^blk\.7\.ffn_gate\.weight$=q2_k
+^blk\.7\.ffn_up\.weight$=q4_k
+^blk\.8\.attn_k\.weight$=q4_k
+^blk\.8\.attn_output\.weight$=q4_k
+^blk\.8\.attn_q\.weight$=q2_k
+^blk\.8\.attn_v\.weight$=q4_k
+^blk\.8\.ffn_down\.weight$=q4_k
+^blk\.8\.ffn_gate\.weight$=q2_k
+^blk\.8\.ffn_up\.weight$=q4_k
+^blk\.9\.attn_k\.weight$=q4_k
+^blk\.9\.attn_output\.weight$=q3_k
+^blk\.9\.attn_q\.weight$=q2_k
+^blk\.9\.attn_v\.weight$=q4_k
+^blk\.9\.ffn_down\.weight$=q2_k
+^blk\.9\.ffn_gate\.weight$=q2_k
+^blk\.9\.ffn_up\.weight$=q2_k
+^blk\.10\.attn_k\.weight$=q4_k
+^blk\.10\.attn_output\.weight$=q3_k
+^blk\.10\.attn_q\.weight$=q2_k
+^blk\.10\.attn_v\.weight$=q4_k
+^blk\.10\.ffn_down\.weight$=q2_k
+^blk\.10\.ffn_gate\.weight$=q2_k
+^blk\.10\.ffn_up\.weight$=q2_k
+^blk\.11\.attn_k\.weight$=q4_k
+^blk\.11\.attn_output\.weight$=q3_k
+^blk\.11\.attn_q\.weight$=q2_k
+^blk\.11\.attn_v\.weight$=q4_k
+^blk\.11\.ffn_down\.weight$=q2_k
+^blk\.11\.ffn_gate\.weight$=q2_k
+^blk\.11\.ffn_up\.weight$=q2_k
+^blk\.12\.attn_k\.weight$=q4_k
+^blk\.12\.attn_output\.weight$=q3_k
+^blk\.12\.attn_q\.weight$=q2_k
+^blk\.12\.attn_v\.weight$=q4_k
+^blk\.12\.ffn_down\.weight$=q2_k
+^blk\.12\.ffn_gate\.weight$=q2_k
+^blk\.12\.ffn_up\.weight$=q2_k
+^blk\.13\.attn_k\.weight$=q4_k
+^blk\.13\.attn_output\.weight$=q3_k
+^blk\.13\.attn_q\.weight$=q2_k
+^blk\.13\.attn_v\.weight$=q4_k
+^blk\.13\.ffn_down\.weight$=q2_k
+^blk\.13\.ffn_gate\.weight$=q2_k
+^blk\.13\.ffn_up\.weight$=q2_k
+^blk\.14\.attn_k\.weight$=q4_k
+^blk\.14\.attn_output\.weight$=q3_k
+^blk\.14\.attn_q\.weight$=q2_k
+^blk\.14\.attn_v\.weight$=q4_k
+^blk\.14\.ffn_down\.weight$=q2_k
+^blk\.14\.ffn_gate\.weight$=q2_k
+^blk\.14\.ffn_up\.weight$=q2_k
+^blk\.15\.attn_k\.weight$=q4_k
+^blk\.15\.attn_output\.weight$=q2_k
+^blk\.15\.attn_q\.weight$=q2_k
+^blk\.15\.attn_v\.weight$=q4_k
+^blk\.15\.ffn_down\.weight$=q2_k
+^blk\.15\.ffn_gate\.weight$=q2_k
+^blk\.15\.ffn_up\.weight$=q2_k
+^blk\.16\.attn_k\.weight$=q4_k
+^blk\.16\.attn_output\.weight$=q2_k
+^blk\.16\.attn_q\.weight$=q2_k
+^blk\.16\.attn_v\.weight$=q4_k
+^blk\.16\.ffn_down\.weight$=q2_k
+^blk\.16\.ffn_gate\.weight$=q2_k
+^blk\.16\.ffn_up\.weight$=q2_k
+^blk\.17\.attn_k\.weight$=q4_k
+^blk\.17\.attn_output\.weight$=q2_k
+^blk\.17\.attn_q\.weight$=q2_k
+^blk\.17\.attn_v\.weight$=q4_k
+^blk\.17\.ffn_down\.weight$=q2_k
+^blk\.17\.ffn_gate\.weight$=q2_k
+^blk\.17\.ffn_up\.weight$=q2_k
+^blk\.18\.attn_k\.weight$=q4_k
+^blk\.18\.attn_output\.weight$=q2_k
+^blk\.18\.attn_q\.weight$=q2_k
+^blk\.18\.attn_v\.weight$=q4_k
+^blk\.18\.ffn_down\.weight$=q2_k
+^blk\.18\.ffn_gate\.weight$=q2_k
+^blk\.18\.ffn_up\.weight$=q2_k
+^blk\.19\.attn_k\.weight$=q4_k
+^blk\.19\.attn_output\.weight$=q2_k
+^blk\.19\.attn_q\.weight$=q2_k
+^blk\.19\.attn_v\.weight$=q4_k
+^blk\.19\.ffn_down\.weight$=q2_k
+^blk\.19\.ffn_gate\.weight$=q2_k
+^blk\.19\.ffn_up\.weight$=q2_k
+^blk\.20\.attn_k\.weight$=q4_k
+^blk\.20\.attn_output\.weight$=q2_k
+^blk\.20\.attn_q\.weight$=q2_k
+^blk\.20\.attn_v\.weight$=q4_k
+^blk\.20\.ffn_down\.weight$=q2_k
+^blk\.20\.ffn_gate\.weight$=q2_k
+^blk\.20\.ffn_up\.weight$=q2_k
+^blk\.21\.attn_k\.weight$=q4_k
+^blk\.21\.attn_output\.weight$=q2_k
+^blk\.21\.attn_q\.weight$=q2_k
+^blk\.21\.attn_v\.weight$=q4_k
+^blk\.21\.ffn_down\.weight$=q2_k
+^blk\.21\.ffn_gate\.weight$=q2_k
+^blk\.21\.ffn_up\.weight$=q2_k
+^blk\.22\.attn_k\.weight$=q4_k
+^blk\.22\.attn_output\.weight$=q2_k
+^blk\.22\.attn_q\.weight$=q2_k
+^blk\.22\.attn_v\.weight$=q4_k
+^blk\.22\.ffn_down\.weight$=q2_k
+^blk\.22\.ffn_gate\.weight$=q2_k
+^blk\.22\.ffn_up\.weight$=q2_k
+^blk\.23\.attn_k\.weight$=q4_k
+^blk\.23\.attn_output\.weight$=q2_k
+^blk\.23\.attn_q\.weight$=q2_k
+^blk\.23\.attn_v\.weight$=q4_k
+^blk\.23\.ffn_down\.weight$=q2_k
+^blk\.23\.ffn_gate\.weight$=q2_k
+^blk\.23\.ffn_up\.weight$=q2_k
+^blk\.24\.attn_k\.weight$=q4_k
+^blk\.24\.attn_output\.weight$=q2_k
+^blk\.24\.attn_q\.weight$=q2_k
+^blk\.24\.attn_v\.weight$=q4_k
+^blk\.24\.ffn_down\.weight$=q2_k
+^blk\.24\.ffn_gate\.weight$=q2_k
+^blk\.24\.ffn_up\.weight$=q2_k
+^blk\.25\.attn_k\.weight$=q4_k
+^blk\.25\.attn_output\.weight$=q2_k
+^blk\.25\.attn_q\.weight$=q2_k
+^blk\.25\.attn_v\.weight$=q4_k
+^blk\.25\.ffn_down\.weight$=q2_k
+^blk\.25\.ffn_gate\.weight$=q2_k
+^blk\.25\.ffn_up\.weight$=q2_k
+^blk\.26\.attn_k\.weight$=q4_k
+^blk\.26\.attn_output\.weight$=q2_k
+^blk\.26\.attn_q\.weight$=q2_k
+^blk\.26\.attn_v\.weight$=q4_k
+^blk\.26\.ffn_down\.weight$=q2_k
+^blk\.26\.ffn_gate\.weight$=q2_k
+^blk\.26\.ffn_up\.weight$=q2_k
+^blk\.27\.attn_k\.weight$=q2_k
+^blk\.27\.attn_output\.weight$=q2_k
+^blk\.27\.attn_q\.weight$=q2_k
+^blk\.27\.attn_v\.weight$=q4_k
+^blk\.27\.ffn_down\.weight$=q2_k
+^blk\.27\.ffn_gate\.weight$=q2_k
+^blk\.27\.ffn_up\.weight$=q2_k
+^blk\.28\.attn_k\.weight$=q2_k
+^blk\.28\.attn_output\.weight$=q2_k
+^blk\.28\.attn_q\.weight$=q2_k
+^blk\.28\.attn_v\.weight$=q4_k
+^blk\.28\.ffn_down\.weight$=q2_k
+^blk\.28\.ffn_gate\.weight$=q2_k
+^blk\.28\.ffn_up\.weight$=q2_k
+^blk\.29\.attn_k\.weight$=q2_k
+^blk\.29\.attn_output\.weight$=q2_k
+^blk\.29\.attn_q\.weight$=q2_k
+^blk\.29\.attn_v\.weight$=q4_k
+^blk\.29\.ffn_down\.weight$=q2_k
+^blk\.29\.ffn_gate\.weight$=q2_k
+^blk\.29\.ffn_up\.weight$=q2_k
+^blk\.30\.attn_k\.weight$=q2_k
+^blk\.30\.attn_output\.weight$=q2_k
+^blk\.30\.attn_q\.weight$=q2_k
+^blk\.30\.attn_v\.weight$=q4_k
+^blk\.30\.ffn_down\.weight$=q2_k
+^blk\.30\.ffn_gate\.weight$=q2_k
+^blk\.30\.ffn_up\.weight$=q2_k
+^blk\.31\.attn_k\.weight$=q2_k
+^blk\.31\.attn_output\.weight$=q2_k
+^blk\.31\.attn_q\.weight$=q2_k
+^blk\.31\.attn_v\.weight$=q4_k
+^blk\.31\.ffn_down\.weight$=q2_k
+^blk\.31\.ffn_gate\.weight$=q2_k
+^blk\.31\.ffn_up\.weight$=q2_k
+^blk\.32\.attn_k\.weight$=q2_k
+^blk\.32\.attn_output\.weight$=q2_k
+^blk\.32\.attn_q\.weight$=q2_k
+^blk\.32\.attn_v\.weight$=q4_k
+^blk\.32\.ffn_down\.weight$=q2_k
+^blk\.32\.ffn_gate\.weight$=q2_k
+^blk\.32\.ffn_up\.weight$=q2_k
+^blk\.33\.attn_k\.weight$=q2_k
+^blk\.33\.attn_output\.weight$=q2_k
+^blk\.33\.attn_q\.weight$=q2_k
+^blk\.33\.attn_v\.weight$=q4_k
+^blk\.33\.ffn_down\.weight$=q2_k
+^blk\.33\.ffn_gate\.weight$=q2_k
+^blk\.33\.ffn_up\.weight$=q2_k
+^blk\.34\.attn_k\.weight$=q2_k
+^blk\.34\.attn_output\.weight$=q2_k
+^blk\.34\.attn_q\.weight$=q2_k
+^blk\.34\.attn_v\.weight$=q4_k
+^blk\.34\.ffn_down\.weight$=q2_k
+^blk\.34\.ffn_gate\.weight$=q2_k
+^blk\.34\.ffn_up\.weight$=q2_k
+^blk\.35\.attn_k\.weight$=q2_k
+^blk\.35\.attn_output\.weight$=q2_k
+^blk\.35\.attn_q\.weight$=q2_k
+^blk\.35\.attn_v\.weight$=q4_k
+^blk\.35\.ffn_down\.weight$=q4_k
+^blk\.35\.ffn_gate\.weight$=q2_k
+^blk\.35\.ffn_up\.weight$=q4_k

layouts/AesCode-8B-Q3_K_L.layout.json

@@ -0,0 +1,88 @@
+{
+  "model": "/media/m2teamgroup/models/AesCode-8B-GGUF/AesCode-8B-bf16.gguf",
+  "imatrix": "/media/m2teamgroup/models/AesCode-8B-GGUF/AesCode-8B-imatrix.gguf",
+  "ftype": "Q3_K_L",
+  "base_type": "q3_k",
+  "family": "k",
+  "recipe": "ident",
+  "variant": "shr0.50",
+  "auto_crush": false,
+  "rung_mode": "share",
+  "tier": "L",
+  "rung_share": 0.5,
+  "crush_pass_bytes": null,
+  "embd_shares": {
+    "output.weight": {
+      "share": 0.1143,
+      "type": "q5_k"
+    },
+    "token_embd.weight": {
+      "share": 0.0599,
+      "type": "q4_k"
+    }
+  },
+  "arch": "dense",
+  "tied": false,
+  "n_layer": 36,
+  "body_rule": true,
+  "pins": {},
+  "shape_pins": {},
+  "fixed_pins": {},
+  "embeddings": {
+    "output.weight": "q5_k",
+    "token_embd.weight": "q4_k"
+  },
+  "bytes": {
+    "heuristic_body+embd": 4466671616,
+    "map_body+embd": 4223500288,
+    "ratio": 0.9456,
+    "note": "quantizable, unpinned tensors only; pinned/unquantized tensors are identical in both"
+  },
+  "body_base_share": 0.5,
+  "body_bpw": 3.969,
+  "body_bpw_heuristic": 4.249,
+  "total_params": 8190735360,
+  "file_bytes_pred": 4224733184,
+  "file_bpw_pred": 4.126,
+  "histogram": {
+    "attn_k": {
+      "q5_k": 27,
+      "q4_k": 9
+    },
+    "attn_output": {
+      "q4_k": 20,
+      "q5_k": 8,
+      "q3_k": 8
+    },
+    "attn_q": {
+      "q3_k": 36
+    },
+    "attn_v": {
+      "q6_k": 27,
+      "q5_k": 9
+    },
+    "ffn_down": {
+      "q4_k": 18,
+      "q3_k": 18
+    },
+    "ffn_gate": {
+      "q3_k": 29,
+      "q4_k": 7
+    },
+    "ffn_up": {
+      "q3_k": 18,
+      "q4_k": 17,
+      "q5_k": 1
+    }
+  },
+  "prior": "layout/prior.json",
+  "pin_rule": "floor",
+  "embd_cap": "heur",
+  "generator": "auto_quant_v2.layout",
+  "generator_key": "3bf8b43e20a68f0b",
+  "quant_type": "Q3_K_L",
+  "quantize_ftype": "Q3_K_L",
+  "llama_cpp_version": "b11490",
+  "unknown_kinds": {},
+  "commit": "5d550ef5561a30307f1d436db2c6835b75395e77"
+}

layouts/AesCode-8B-Q3_K_L.tensor-types.txt

@@ -0,0 +1,254 @@
+^output\.weight$=q5_k
+^token_embd\.weight$=q4_k
+^blk\.0\.attn_k\.weight$=q5_k
+^blk\.0\.attn_output\.weight$=q4_k
+^blk\.0\.attn_q\.weight$=q3_k
+^blk\.0\.attn_v\.weight$=q5_k
+^blk\.0\.ffn_down\.weight$=q4_k
+^blk\.0\.ffn_gate\.weight$=q3_k
+^blk\.0\.ffn_up\.weight$=q4_k
+^blk\.1\.attn_k\.weight$=q5_k
+^blk\.1\.attn_output\.weight$=q5_k
+^blk\.1\.attn_q\.weight$=q3_k
+^blk\.1\.attn_v\.weight$=q6_k
+^blk\.1\.ffn_down\.weight$=q4_k
+^blk\.1\.ffn_gate\.weight$=q4_k
+^blk\.1\.ffn_up\.weight$=q4_k
+^blk\.2\.attn_k\.weight$=q5_k
+^blk\.2\.attn_output\.weight$=q5_k
+^blk\.2\.attn_q\.weight$=q3_k
+^blk\.2\.attn_v\.weight$=q6_k
+^blk\.2\.ffn_down\.weight$=q4_k
+^blk\.2\.ffn_gate\.weight$=q4_k
+^blk\.2\.ffn_up\.weight$=q4_k
+^blk\.3\.attn_k\.weight$=q5_k
+^blk\.3\.attn_output\.weight$=q5_k
+^blk\.3\.attn_q\.weight$=q3_k
+^blk\.3\.attn_v\.weight$=q6_k
+^blk\.3\.ffn_down\.weight$=q4_k
+^blk\.3\.ffn_gate\.weight$=q4_k
+^blk\.3\.ffn_up\.weight$=q4_k
+^blk\.4\.attn_k\.weight$=q5_k
+^blk\.4\.attn_output\.weight$=q5_k
+^blk\.4\.attn_q\.weight$=q3_k
+^blk\.4\.attn_v\.weight$=q6_k
+^blk\.4\.ffn_down\.weight$=q4_k
+^blk\.4\.ffn_gate\.weight$=q4_k
+^blk\.4\.ffn_up\.weight$=q4_k
+^blk\.5\.attn_k\.weight$=q5_k
+^blk\.5\.attn_output\.weight$=q5_k
+^blk\.5\.attn_q\.weight$=q3_k
+^blk\.5\.attn_v\.weight$=q6_k
+^blk\.5\.ffn_down\.weight$=q4_k
+^blk\.5\.ffn_gate\.weight$=q4_k
+^blk\.5\.ffn_up\.weight$=q4_k
+^blk\.6\.attn_k\.weight$=q5_k
+^blk\.6\.attn_output\.weight$=q5_k
+^blk\.6\.attn_q\.weight$=q3_k
+^blk\.6\.attn_v\.weight$=q6_k
+^blk\.6\.ffn_down\.weight$=q4_k
+^blk\.6\.ffn_gate\.weight$=q4_k
+^blk\.6\.ffn_up\.weight$=q4_k
+^blk\.7\.attn_k\.weight$=q5_k
+^blk\.7\.attn_output\.weight$=q5_k
+^blk\.7\.attn_q\.weight$=q3_k
+^blk\.7\.attn_v\.weight$=q6_k
+^blk\.7\.ffn_down\.weight$=q4_k
+^blk\.7\.ffn_gate\.weight$=q3_k
+^blk\.7\.ffn_up\.weight$=q4_k
+^blk\.8\.attn_k\.weight$=q5_k
+^blk\.8\.attn_output\.weight$=q5_k
+^blk\.8\.attn_q\.weight$=q3_k
+^blk\.8\.attn_v\.weight$=q6_k
+^blk\.8\.ffn_down\.weight$=q4_k
+^blk\.8\.ffn_gate\.weight$=q3_k
+^blk\.8\.ffn_up\.weight$=q4_k
+^blk\.9\.attn_k\.weight$=q5_k
+^blk\.9\.attn_output\.weight$=q4_k
+^blk\.9\.attn_q\.weight$=q3_k
+^blk\.9\.attn_v\.weight$=q6_k
+^blk\.9\.ffn_down\.weight$=q3_k
+^blk\.9\.ffn_gate\.weight$=q3_k
+^blk\.9\.ffn_up\.weight$=q3_k
+^blk\.10\.attn_k\.weight$=q5_k
+^blk\.10\.attn_output\.weight$=q4_k
+^blk\.10\.attn_q\.weight$=q3_k
+^blk\.10\.attn_v\.weight$=q6_k
+^blk\.10\.ffn_down\.weight$=q3_k
+^blk\.10\.ffn_gate\.weight$=q3_k
+^blk\.10\.ffn_up\.weight$=q3_k
+^blk\.11\.attn_k\.weight$=q5_k
+^blk\.11\.attn_output\.weight$=q4_k
+^blk\.11\.attn_q\.weight$=q3_k
+^blk\.11\.attn_v\.weight$=q6_k
+^blk\.11\.ffn_down\.weight$=q3_k
+^blk\.11\.ffn_gate\.weight$=q3_k
+^blk\.11\.ffn_up\.weight$=q3_k
+^blk\.12\.attn_k\.weight$=q5_k
+^blk\.12\.attn_output\.weight$=q4_k
+^blk\.12\.attn_q\.weight$=q3_k
+^blk\.12\.attn_v\.weight$=q6_k
+^blk\.12\.ffn_down\.weight$=q3_k
+^blk\.12\.ffn_gate\.weight$=q3_k
+^blk\.12\.ffn_up\.weight$=q3_k
+^blk\.13\.attn_k\.weight$=q5_k
+^blk\.13\.attn_output\.weight$=q4_k
+^blk\.13\.attn_q\.weight$=q3_k
+^blk\.13\.attn_v\.weight$=q6_k
+^blk\.13\.ffn_down\.weight$=q3_k
+^blk\.13\.ffn_gate\.weight$=q3_k
+^blk\.13\.ffn_up\.weight$=q3_k
+^blk\.14\.attn_k\.weight$=q5_k
+^blk\.14\.attn_output\.weight$=q4_k
+^blk\.14\.attn_q\.weight$=q3_k
+^blk\.14\.attn_v\.weight$=q6_k
+^blk\.14\.ffn_down\.weight$=q3_k
+^blk\.14\.ffn_gate\.weight$=q3_k
+^blk\.14\.ffn_up\.weight$=q3_k
+^blk\.15\.attn_k\.weight$=q5_k
+^blk\.15\.attn_output\.weight$=q4_k
+^blk\.15\.attn_q\.weight$=q3_k
+^blk\.15\.attn_v\.weight$=q6_k
+^blk\.15\.ffn_down\.weight$=q3_k
+^blk\.15\.ffn_gate\.weight$=q3_k
+^blk\.15\.ffn_up\.weight$=q3_k
+^blk\.16\.attn_k\.weight$=q5_k
+^blk\.16\.attn_output\.weight$=q4_k
+^blk\.16\.attn_q\.weight$=q3_k
+^blk\.16\.attn_v\.weight$=q6_k
+^blk\.16\.ffn_down\.weight$=q3_k
+^blk\.16\.ffn_gate\.weight$=q3_k
+^blk\.16\.ffn_up\.weight$=q3_k
+^blk\.17\.attn_k\.weight$=q5_k
+^blk\.17\.attn_output\.weight$=q4_k
+^blk\.17\.attn_q\.weight$=q3_k
+^blk\.17\.attn_v\.weight$=q6_k
+^blk\.17\.ffn_down\.weight$=q3_k
+^blk\.17\.ffn_gate\.weight$=q3_k
+^blk\.17\.ffn_up\.weight$=q3_k
+^blk\.18\.attn_k\.weight$=q5_k
+^blk\.18\.attn_output\.weight$=q4_k
+^blk\.18\.attn_q\.weight$=q3_k
+^blk\.18\.attn_v\.weight$=q6_k
+^blk\.18\.ffn_down\.weight$=q3_k
+^blk\.18\.ffn_gate\.weight$=q3_k
+^blk\.18\.ffn_up\.weight$=q3_k
+^blk\.19\.attn_k\.weight$=q5_k
+^blk\.19\.attn_output\.weight$=q4_k
+^blk\.19\.attn_q\.weight$=q3_k
+^blk\.19\.attn_v\.weight$=q6_k
+^blk\.19\.ffn_down\.weight$=q3_k
+^blk\.19\.ffn_gate\.weight$=q3_k
+^blk\.19\.ffn_up\.weight$=q3_k
+^blk\.20\.attn_k\.weight$=q5_k
+^blk\.20\.attn_output\.weight$=q4_k
+^blk\.20\.attn_q\.weight$=q3_k
+^blk\.20\.attn_v\.weight$=q6_k
+^blk\.20\.ffn_down\.weight$=q3_k
+^blk\.20\.ffn_gate\.weight$=q3_k
+^blk\.20\.ffn_up\.weight$=q3_k
+^blk\.21\.attn_k\.weight$=q5_k
+^blk\.21\.attn_output\.weight$=q4_k
+^blk\.21\.attn_q\.weight$=q3_k
+^blk\.21\.attn_v\.weight$=q6_k
+^blk\.21\.ffn_down\.weight$=q3_k
+^blk\.21\.ffn_gate\.weight$=q3_k
+^blk\.21\.ffn_up\.weight$=q3_k
+^blk\.22\.attn_k\.weight$=q5_k
+^blk\.22\.attn_output\.weight$=q4_k
+^blk\.22\.attn_q\.weight$=q3_k
+^blk\.22\.attn_v\.weight$=q6_k
+^blk\.22\.ffn_down\.weight$=q3_k
+^blk\.22\.ffn_gate\.weight$=q3_k
+^blk\.22\.ffn_up\.weight$=q3_k
+^blk\.23\.attn_k\.weight$=q5_k
+^blk\.23\.attn_output\.weight$=q4_k
+^blk\.23\.attn_q\.weight$=q3_k
+^blk\.23\.attn_v\.weight$=q6_k
+^blk\.23\.ffn_down\.weight$=q3_k
+^blk\.23\.ffn_gate\.weight$=q3_k
+^blk\.23\.ffn_up\.weight$=q3_k
+^blk\.24\.attn_k\.weight$=q5_k
+^blk\.24\.attn_output\.weight$=q4_k
+^blk\.24\.attn_q\.weight$=q3_k
+^blk\.24\.attn_v\.weight$=q6_k
+^blk\.24\.ffn_down\.weight$=q3_k
+^blk\.24\.ffn_gate\.weight$=q3_k
+^blk\.24\.ffn_up\.weight$=q3_k
+^blk\.25\.attn_k\.weight$=q5_k
+^blk\.25\.attn_output\.weight$=q4_k
+^blk\.25\.attn_q\.weight$=q3_k
+^blk\.25\.attn_v\.weight$=q6_k
+^blk\.25\.ffn_down\.weight$=q3_k
+^blk\.25\.ffn_gate\.weight$=q3_k
+^blk\.25\.ffn_up\.weight$=q3_k
+^blk\.26\.attn_k\.weight$=q5_k
+^blk\.26\.attn_output\.weight$=q4_k
+^blk\.26\.attn_q\.weight$=q3_k
+^blk\.26\.attn_v\.weight$=q6_k
+^blk\.26\.ffn_down\.weight$=q3_k
+^blk\.26\.ffn_gate\.weight$=q3_k
+^blk\.26\.ffn_up\.weight$=q3_k
+^blk\.27\.attn_k\.weight$=q4_k
+^blk\.27\.attn_output\.weight$=q3_k
+^blk\.27\.attn_q\.weight$=q3_k
+^blk\.27\.attn_v\.weight$=q5_k
+^blk\.27\.ffn_down\.weight$=q4_k
+^blk\.27\.ffn_gate\.weight$=q3_k
+^blk\.27\.ffn_up\.weight$=q4_k
+^blk\.28\.attn_k\.weight$=q4_k
+^blk\.28\.attn_output\.weight$=q3_k
+^blk\.28\.attn_q\.weight$=q3_k
+^blk\.28\.attn_v\.weight$=q5_k
+^blk\.28\.ffn_down\.weight$=q4_k
+^blk\.28\.ffn_gate\.weight$=q3_k
+^blk\.28\.ffn_up\.weight$=q4_k
+^blk\.29\.attn_k\.weight$=q4_k
+^blk\.29\.attn_output\.weight$=q3_k
+^blk\.29\.attn_q\.weight$=q3_k
+^blk\.29\.attn_v\.weight$=q5_k
+^blk\.29\.ffn_down\.weight$=q4_k
+^blk\.29\.ffn_gate\.weight$=q3_k
+^blk\.29\.ffn_up\.weight$=q4_k
+^blk\.30\.attn_k\.weight$=q4_k
+^blk\.30\.attn_output\.weight$=q3_k
+^blk\.30\.attn_q\.weight$=q3_k
+^blk\.30\.attn_v\.weight$=q5_k
+^blk\.30\.ffn_down\.weight$=q4_k
+^blk\.30\.ffn_gate\.weight$=q3_k
+^blk\.30\.ffn_up\.weight$=q4_k
+^blk\.31\.attn_k\.weight$=q4_k
+^blk\.31\.attn_output\.weight$=q3_k
+^blk\.31\.attn_q\.weight$=q3_k
+^blk\.31\.attn_v\.weight$=q5_k
+^blk\.31\.ffn_down\.weight$=q4_k
+^blk\.31\.ffn_gate\.weight$=q3_k
+^blk\.31\.ffn_up\.weight$=q4_k
+^blk\.32\.attn_k\.weight$=q4_k
+^blk\.32\.attn_output\.weight$=q3_k
+^blk\.32\.attn_q\.weight$=q3_k
+^blk\.32\.attn_v\.weight$=q5_k
+^blk\.32\.ffn_down\.weight$=q4_k
+^blk\.32\.ffn_gate\.weight$=q3_k
+^blk\.32\.ffn_up\.weight$=q4_k
+^blk\.33\.attn_k\.weight$=q4_k
+^blk\.33\.attn_output\.weight$=q3_k
+^blk\.33\.attn_q\.weight$=q3_k
+^blk\.33\.attn_v\.weight$=q5_k
+^blk\.33\.ffn_down\.weight$=q4_k
+^blk\.33\.ffn_gate\.weight$=q3_k
+^blk\.33\.ffn_up\.weight$=q4_k
+^blk\.34\.attn_k\.weight$=q4_k
+^blk\.34\.attn_output\.weight$=q3_k
+^blk\.34\.attn_q\.weight$=q3_k
+^blk\.34\.attn_v\.weight$=q5_k
+^blk\.34\.ffn_down\.weight$=q4_k
+^blk\.34\.ffn_gate\.weight$=q3_k
+^blk\.34\.ffn_up\.weight$=q4_k
+^blk\.35\.attn_k\.weight$=q4_k
+^blk\.35\.attn_output\.weight$=q4_k
+^blk\.35\.attn_q\.weight$=q3_k
+^blk\.35\.attn_v\.weight$=q6_k
+^blk\.35\.ffn_down\.weight$=q4_k
+^blk\.35\.ffn_gate\.weight$=q4_k
+^blk\.35\.ffn_up\.weight$=q5_k

layouts/AesCode-8B-Q3_K_M.layout.json

@@ -0,0 +1,86 @@
+{
+  "model": "/media/m2teamgroup/models/AesCode-8B-GGUF/AesCode-8B-bf16.gguf",
+  "imatrix": "/media/m2teamgroup/models/AesCode-8B-GGUF/AesCode-8B-imatrix.gguf",
+  "ftype": "Q3_K_M",
+  "base_type": "q3_k",
+  "family": "k",
+  "recipe": "ident",
+  "variant": "shr0.70",
+  "auto_crush": false,
+  "rung_mode": "share",
+  "tier": "M",
+  "rung_share": 0.7,
+  "crush_pass_bytes": null,
+  "embd_shares": {
+    "output.weight": {
+      "share": 0.1203,
+      "type": "q5_k"
+    },
+    "token_embd.weight": {
+      "share": 0.063,
+      "type": "q4_k"
+    }
+  },
+  "arch": "dense",
+  "tied": false,
+  "n_layer": 36,
+  "body_rule": true,
+  "pins": {},
+  "shape_pins": {},
+  "fixed_pins": {},
+  "embeddings": {
+    "output.weight": "q5_k",
+    "token_embd.weight": "q4_k"
+  },
+  "bytes": {
+    "heuristic_body+embd": 4243324928,
+    "map_body+embd": 4032233472,
+    "ratio": 0.9503,
+    "note": "quantizable, unpinned tensors only; pinned/unquantized tensors are identical in both"
+  },
+  "body_base_share": 0.7,
+  "body_bpw": 3.748,
+  "body_bpw_heuristic": 3.991,
+  "total_params": 8190735360,
+  "file_bytes_pred": 4033466368,
+  "file_bpw_pred": 3.94,
+  "histogram": {
+    "attn_k": {
+      "q5_k": 28,
+      "q4_k": 8
+    },
+    "attn_output": {
+      "q4_k": 18,
+      "q3_k": 10,
+      "q5_k": 8
+    },
+    "attn_q": {
+      "q3_k": 36
+    },
+    "attn_v": {
+      "q5_k": 34,
+      "q6_k": 2
+    },
+    "ffn_down": {
+      "q3_k": 27,
+      "q4_k": 9
+    },
+    "ffn_gate": {
+      "q3_k": 36
+    },
+    "ffn_up": {
+      "q3_k": 27,
+      "q4_k": 9
+    }
+  },
+  "prior": "layout/prior.json",
+  "pin_rule": "floor",
+  "embd_cap": "heur",
+  "generator": "auto_quant_v2.layout",
+  "generator_key": "3bf8b43e20a68f0b",
+  "quant_type": "Q3_K_M",
+  "quantize_ftype": "Q3_K_M",
+  "llama_cpp_version": "b11490",
+  "unknown_kinds": {},
+  "commit": "5d550ef5561a30307f1d436db2c6835b75395e77"
+}

layouts/AesCode-8B-Q3_K_M.tensor-types.txt

@@ -0,0 +1,254 @@
+^output\.weight$=q5_k
+^token_embd\.weight$=q4_k
+^blk\.0\.attn_k\.weight$=q5_k
+^blk\.0\.attn_output\.weight$=q3_k
+^blk\.0\.attn_q\.weight$=q3_k
+^blk\.0\.attn_v\.weight$=q5_k
+^blk\.0\.ffn_down\.weight$=q3_k
+^blk\.0\.ffn_gate\.weight$=q3_k
+^blk\.0\.ffn_up\.weight$=q3_k
+^blk\.1\.attn_k\.weight$=q5_k
+^blk\.1\.attn_output\.weight$=q5_k
+^blk\.1\.attn_q\.weight$=q3_k
+^blk\.1\.attn_v\.weight$=q6_k
+^blk\.1\.ffn_down\.weight$=q4_k
+^blk\.1\.ffn_gate\.weight$=q3_k
+^blk\.1\.ffn_up\.weight$=q4_k
+^blk\.2\.attn_k\.weight$=q5_k
+^blk\.2\.attn_output\.weight$=q5_k
+^blk\.2\.attn_q\.weight$=q3_k
+^blk\.2\.attn_v\.weight$=q6_k
+^blk\.2\.ffn_down\.weight$=q4_k
+^blk\.2\.ffn_gate\.weight$=q3_k
+^blk\.2\.ffn_up\.weight$=q4_k
+^blk\.3\.attn_k\.weight$=q5_k
+^blk\.3\.attn_output\.weight$=q5_k
+^blk\.3\.attn_q\.weight$=q3_k
+^blk\.3\.attn_v\.weight$=q5_k
+^blk\.3\.ffn_down\.weight$=q4_k
+^blk\.3\.ffn_gate\.weight$=q3_k
+^blk\.3\.ffn_up\.weight$=q4_k
+^blk\.4\.attn_k\.weight$=q5_k
+^blk\.4\.attn_output\.weight$=q5_k
+^blk\.4\.attn_q\.weight$=q3_k
+^blk\.4\.attn_v\.weight$=q5_k
+^blk\.4\.ffn_down\.weight$=q4_k
+^blk\.4\.ffn_gate\.weight$=q3_k
+^blk\.4\.ffn_up\.weight$=q4_k
+^blk\.5\.attn_k\.weight$=q5_k
+^blk\.5\.attn_output\.weight$=q5_k
+^blk\.5\.attn_q\.weight$=q3_k
+^blk\.5\.attn_v\.weight$=q5_k
+^blk\.5\.ffn_down\.weight$=q4_k
+^blk\.5\.ffn_gate\.weight$=q3_k
+^blk\.5\.ffn_up\.weight$=q4_k
+^blk\.6\.attn_k\.weight$=q5_k
+^blk\.6\.attn_output\.weight$=q5_k
+^blk\.6\.attn_q\.weight$=q3_k
+^blk\.6\.attn_v\.weight$=q5_k
+^blk\.6\.ffn_down\.weight$=q4_k
+^blk\.6\.ffn_gate\.weight$=q3_k
+^blk\.6\.ffn_up\.weight$=q4_k
+^blk\.7\.attn_k\.weight$=q5_k
+^blk\.7\.attn_output\.weight$=q5_k
+^blk\.7\.attn_q\.weight$=q3_k
+^blk\.7\.attn_v\.weight$=q5_k
+^blk\.7\.ffn_down\.weight$=q4_k
+^blk\.7\.ffn_gate\.weight$=q3_k
+^blk\.7\.ffn_up\.weight$=q4_k
+^blk\.8\.attn_k\.weight$=q5_k
+^blk\.8\.attn_output\.weight$=q5_k
+^blk\.8\.attn_q\.weight$=q3_k
+^blk\.8\.attn_v\.weight$=q5_k
+^blk\.8\.ffn_down\.weight$=q4_k
+^blk\.8\.ffn_gate\.weight$=q3_k
+^blk\.8\.ffn_up\.weight$=q4_k
+^blk\.9\.attn_k\.weight$=q5_k
+^blk\.9\.attn_output\.weight$=q4_k
+^blk\.9\.attn_q\.weight$=q3_k
+^blk\.9\.attn_v\.weight$=q5_k
+^blk\.9\.ffn_down\.weight$=q3_k
+^blk\.9\.ffn_gate\.weight$=q3_k
+^blk\.9\.ffn_up\.weight$=q3_k
+^blk\.10\.attn_k\.weight$=q5_k
+^blk\.10\.attn_output\.weight$=q4_k
+^blk\.10\.attn_q\.weight$=q3_k
+^blk\.10\.attn_v\.weight$=q5_k
+^blk\.10\.ffn_down\.weight$=q3_k
+^blk\.10\.ffn_gate\.weight$=q3_k
+^blk\.10\.ffn_up\.weight$=q3_k
+^blk\.11\.attn_k\.weight$=q5_k
+^blk\.11\.attn_output\.weight$=q4_k
+^blk\.11\.attn_q\.weight$=q3_k
+^blk\.11\.attn_v\.weight$=q5_k
+^blk\.11\.ffn_down\.weight$=q3_k
+^blk\.11\.ffn_gate\.weight$=q3_k
+^blk\.11\.ffn_up\.weight$=q3_k
+^blk\.12\.attn_k\.weight$=q5_k
+^blk\.12\.attn_output\.weight$=q4_k
+^blk\.12\.attn_q\.weight$=q3_k
+^blk\.12\.attn_v\.weight$=q5_k
+^blk\.12\.ffn_down\.weight$=q3_k
+^blk\.12\.ffn_gate\.weight$=q3_k
+^blk\.12\.ffn_up\.weight$=q3_k
+^blk\.13\.attn_k\.weight$=q5_k
+^blk\.13\.attn_output\.weight$=q4_k
+^blk\.13\.attn_q\.weight$=q3_k
+^blk\.13\.attn_v\.weight$=q5_k
+^blk\.13\.ffn_down\.weight$=q3_k
+^blk\.13\.ffn_gate\.weight$=q3_k
+^blk\.13\.ffn_up\.weight$=q3_k
+^blk\.14\.attn_k\.weight$=q5_k
+^blk\.14\.attn_output\.weight$=q4_k
+^blk\.14\.attn_q\.weight$=q3_k
+^blk\.14\.attn_v\.weight$=q5_k
+^blk\.14\.ffn_down\.weight$=q3_k
+^blk\.14\.ffn_gate\.weight$=q3_k
+^blk\.14\.ffn_up\.weight$=q3_k
+^blk\.15\.attn_k\.weight$=q5_k
+^blk\.15\.attn_output\.weight$=q4_k
+^blk\.15\.attn_q\.weight$=q3_k
+^blk\.15\.attn_v\.weight$=q5_k
+^blk\.15\.ffn_down\.weight$=q3_k
+^blk\.15\.ffn_gate\.weight$=q3_k
+^blk\.15\.ffn_up\.weight$=q3_k
+^blk\.16\.attn_k\.weight$=q5_k
+^blk\.16\.attn_output\.weight$=q4_k
+^blk\.16\.attn_q\.weight$=q3_k
+^blk\.16\.attn_v\.weight$=q5_k
+^blk\.16\.ffn_down\.weight$=q3_k
+^blk\.16\.ffn_gate\.weight$=q3_k
+^blk\.16\.ffn_up\.weight$=q3_k
+^blk\.17\.attn_k\.weight$=q5_k
+^blk\.17\.attn_output\.weight$=q4_k
+^blk\.17\.attn_q\.weight$=q3_k
+^blk\.17\.attn_v\.weight$=q5_k
+^blk\.17\.ffn_down\.weight$=q3_k
+^blk\.17\.ffn_gate\.weight$=q3_k
+^blk\.17\.ffn_up\.weight$=q3_k
+^blk\.18\.attn_k\.weight$=q5_k
+^blk\.18\.attn_output\.weight$=q4_k
+^blk\.18\.attn_q\.weight$=q3_k
+^blk\.18\.attn_v\.weight$=q5_k
+^blk\.18\.ffn_down\.weight$=q3_k
+^blk\.18\.ffn_gate\.weight$=q3_k
+^blk\.18\.ffn_up\.weight$=q3_k
+^blk\.19\.attn_k\.weight$=q5_k
+^blk\.19\.attn_output\.weight$=q4_k
+^blk\.19\.attn_q\.weight$=q3_k
+^blk\.19\.attn_v\.weight$=q5_k
+^blk\.19\.ffn_down\.weight$=q3_k
+^blk\.19\.ffn_gate\.weight$=q3_k
+^blk\.19\.ffn_up\.weight$=q3_k
+^blk\.20\.attn_k\.weight$=q5_k
+^blk\.20\.attn_output\.weight$=q4_k
+^blk\.20\.attn_q\.weight$=q3_k
+^blk\.20\.attn_v\.weight$=q5_k
+^blk\.20\.ffn_down\.weight$=q3_k
+^blk\.20\.ffn_gate\.weight$=q3_k
+^blk\.20\.ffn_up\.weight$=q3_k
+^blk\.21\.attn_k\.weight$=q5_k
+^blk\.21\.attn_output\.weight$=q4_k
+^blk\.21\.attn_q\.weight$=q3_k
+^blk\.21\.attn_v\.weight$=q5_k
+^blk\.21\.ffn_down\.weight$=q3_k
+^blk\.21\.ffn_gate\.weight$=q3_k
+^blk\.21\.ffn_up\.weight$=q3_k
+^blk\.22\.attn_k\.weight$=q5_k
+^blk\.22\.attn_output\.weight$=q4_k
+^blk\.22\.attn_q\.weight$=q3_k
+^blk\.22\.attn_v\.weight$=q5_k
+^blk\.22\.ffn_down\.weight$=q3_k
+^blk\.22\.ffn_gate\.weight$=q3_k
+^blk\.22\.ffn_up\.weight$=q3_k
+^blk\.23\.attn_k\.weight$=q5_k
+^blk\.23\.attn_output\.weight$=q4_k
+^blk\.23\.attn_q\.weight$=q3_k
+^blk\.23\.attn_v\.weight$=q5_k
+^blk\.23\.ffn_down\.weight$=q3_k
+^blk\.23\.ffn_gate\.weight$=q3_k
+^blk\.23\.ffn_up\.weight$=q3_k
+^blk\.24\.attn_k\.weight$=q5_k
+^blk\.24\.attn_output\.weight$=q4_k
+^blk\.24\.attn_q\.weight$=q3_k
+^blk\.24\.attn_v\.weight$=q5_k
+^blk\.24\.ffn_down\.weight$=q3_k
+^blk\.24\.ffn_gate\.weight$=q3_k
+^blk\.24\.ffn_up\.weight$=q3_k
+^blk\.25\.attn_k\.weight$=q5_k
+^blk\.25\.attn_output\.weight$=q4_k
+^blk\.25\.attn_q\.weight$=q3_k
+^blk\.25\.attn_v\.weight$=q5_k
+^blk\.25\.ffn_down\.weight$=q3_k
+^blk\.25\.ffn_gate\.weight$=q3_k
+^blk\.25\.ffn_up\.weight$=q3_k
+^blk\.26\.attn_k\.weight$=q5_k
+^blk\.26\.attn_output\.weight$=q4_k
+^blk\.26\.attn_q\.weight$=q3_k
+^blk\.26\.attn_v\.weight$=q5_k
+^blk\.26\.ffn_down\.weight$=q3_k
+^blk\.26\.ffn_gate\.weight$=q3_k
+^blk\.26\.ffn_up\.weight$=q3_k
+^blk\.27\.attn_k\.weight$=q5_k
+^blk\.27\.attn_output\.weight$=q3_k
+^blk\.27\.attn_q\.weight$=q3_k
+^blk\.27\.attn_v\.weight$=q5_k
+^blk\.27\.ffn_down\.weight$=q3_k
+^blk\.27\.ffn_gate\.weight$=q3_k
+^blk\.27\.ffn_up\.weight$=q3_k
+^blk\.28\.attn_k\.weight$=q4_k
+^blk\.28\.attn_output\.weight$=q3_k
+^blk\.28\.attn_q\.weight$=q3_k
+^blk\.28\.attn_v\.weight$=q5_k
+^blk\.28\.ffn_down\.weight$=q3_k
+^blk\.28\.ffn_gate\.weight$=q3_k
+^blk\.28\.ffn_up\.weight$=q3_k
+^blk\.29\.attn_k\.weight$=q4_k
+^blk\.29\.attn_output\.weight$=q3_k
+^blk\.29\.attn_q\.weight$=q3_k
+^blk\.29\.attn_v\.weight$=q5_k
+^blk\.29\.ffn_down\.weight$=q3_k
+^blk\.29\.ffn_gate\.weight$=q3_k
+^blk\.29\.ffn_up\.weight$=q3_k
+^blk\.30\.attn_k\.weight$=q4_k
+^blk\.30\.attn_output\.weight$=q3_k
+^blk\.30\.attn_q\.weight$=q3_k
+^blk\.30\.attn_v\.weight$=q5_k
+^blk\.30\.ffn_down\.weight$=q3_k
+^blk\.30\.ffn_gate\.weight$=q3_k
+^blk\.30\.ffn_up\.weight$=q3_k
+^blk\.31\.attn_k\.weight$=q4_k
+^blk\.31\.attn_output\.weight$=q3_k
+^blk\.31\.attn_q\.weight$=q3_k
+^blk\.31\.attn_v\.weight$=q5_k
+^blk\.31\.ffn_down\.weight$=q3_k
+^blk\.31\.ffn_gate\.weight$=q3_k
+^blk\.31\.ffn_up\.weight$=q3_k
+^blk\.32\.attn_k\.weight$=q4_k
+^blk\.32\.attn_output\.weight$=q3_k
+^blk\.32\.attn_q\.weight$=q3_k
+^blk\.32\.attn_v\.weight$=q5_k
+^blk\.32\.ffn_down\.weight$=q3_k
+^blk\.32\.ffn_gate\.weight$=q3_k
+^blk\.32\.ffn_up\.weight$=q3_k
+^blk\.33\.attn_k\.weight$=q4_k
+^blk\.33\.attn_output\.weight$=q3_k
+^blk\.33\.attn_q\.weight$=q3_k
+^blk\.33\.attn_v\.weight$=q5_k
+^blk\.33\.ffn_down\.weight$=q3_k
+^blk\.33\.ffn_gate\.weight$=q3_k
+^blk\.33\.ffn_up\.weight$=q3_k
+^blk\.34\.attn_k\.weight$=q4_k
+^blk\.34\.attn_output\.weight$=q3_k
+^blk\.34\.attn_q\.weight$=q3_k
+^blk\.34\.attn_v\.weight$=q5_k
+^blk\.34\.ffn_down\.weight$=q3_k
+^blk\.34\.ffn_gate\.weight$=q3_k
+^blk\.34\.ffn_up\.weight$=q3_k
+^blk\.35\.attn_k\.weight$=q4_k
+^blk\.35\.attn_output\.weight$=q3_k
+^blk\.35\.attn_q\.weight$=q3_k
+^blk\.35\.attn_v\.weight$=q5_k
+^blk\.35\.ffn_down\.weight$=q4_k
+^blk\.35\.ffn_gate\.weight$=q3_k
+^blk\.35\.ffn_up\.weight$=q4_k

layouts/AesCode-8B-Q3_K_S.layout.json

@@ -0,0 +1,85 @@
+{
+  "model": "/media/m2teamgroup/models/AesCode-8B-GGUF/AesCode-8B-bf16.gguf",
+  "imatrix": "/media/m2teamgroup/models/AesCode-8B-GGUF/AesCode-8B-imatrix.gguf",
+  "ftype": "Q3_K_S",
+  "base_type": "q3_k",
+  "family": "k",
+  "recipe": "ident",
+  "variant": "shr0.90",
+  "auto_crush": false,
+  "rung_mode": "share",
+  "tier": "S",
+  "rung_share": 0.9,
+  "crush_pass_bytes": null,
+  "embd_shares": {
+    "output.weight": {
+      "share": 0.1315,
+      "type": "q5_k"
+    },
+    "token_embd.weight": {
+      "share": 0.0689,
+      "type": "q4_k"
+    }
+  },
+  "arch": "dense",
+  "tied": false,
+  "n_layer": 36,
+  "body_rule": true,
+  "pins": {},
+  "shape_pins": {},
+  "fixed_pins": {},
+  "embeddings": {
+    "output.weight": "q5_k",
+    "token_embd.weight": "q4_k"
+  },
+  "bytes": {
+    "heuristic_body+embd": 3880386560,
+    "map_body+embd": 3863707648,
+    "ratio": 0.9957,
+    "note": "quantizable, unpinned tensors only; pinned/unquantized tensors are identical in both"
+  },
+  "body_base_share": 0.9,
+  "body_bpw": 3.554,
+  "body_bpw_heuristic": 3.573,
+  "total_params": 8190735360,
+  "file_bytes_pred": 3864940544,
+  "file_bpw_pred": 3.775,
+  "histogram": {
+    "attn_k": {
+      "q4_k": 24,
+      "q3_k": 9,
+      "q5_k": 3
+    },
+    "attn_output": {
+      "q3_k": 26,
+      "q5_k": 8,
+      "q4_k": 2
+    },
+    "attn_q": {
+      "q3_k": 36
+    },
+    "attn_v": {
+      "q5_k": 36
+    },
+    "ffn_down": {
+      "q3_k": 36
+    },
+    "ffn_gate": {
+      "q3_k": 36
+    },
+    "ffn_up": {
+      "q3_k": 35,
+      "q4_k": 1
+    }
+  },
+  "prior": "layout/prior.json",
+  "pin_rule": "floor",
+  "embd_cap": "heur",
+  "generator": "auto_quant_v2.layout",
+  "generator_key": "3bf8b43e20a68f0b",
+  "quant_type": "Q3_K_S",
+  "quantize_ftype": "Q3_K_S",
+  "llama_cpp_version": "b11490",
+  "unknown_kinds": {},
+  "commit": "5d550ef5561a30307f1d436db2c6835b75395e77"
+}

layouts/AesCode-8B-Q3_K_S.tensor-types.txt

@@ -0,0 +1,254 @@
+^output\.weight$=q5_k
+^token_embd\.weight$=q4_k
+^blk\.0\.attn_k\.weight$=q5_k
+^blk\.0\.attn_output\.weight$=q3_k
+^blk\.0\.attn_q\.weight$=q3_k
+^blk\.0\.attn_v\.weight$=q5_k
+^blk\.0\.ffn_down\.weight$=q3_k
+^blk\.0\.ffn_gate\.weight$=q3_k
+^blk\.0\.ffn_up\.weight$=q3_k
+^blk\.1\.attn_k\.weight$=q4_k
+^blk\.1\.attn_output\.weight$=q5_k
+^blk\.1\.attn_q\.weight$=q3_k
+^blk\.1\.attn_v\.weight$=q5_k
+^blk\.1\.ffn_down\.weight$=q3_k
+^blk\.1\.ffn_gate\.weight$=q3_k
+^blk\.1\.ffn_up\.weight$=q3_k
+^blk\.2\.attn_k\.weight$=q4_k
+^blk\.2\.attn_output\.weight$=q5_k
+^blk\.2\.attn_q\.weight$=q3_k
+^blk\.2\.attn_v\.weight$=q5_k
+^blk\.2\.ffn_down\.weight$=q3_k
+^blk\.2\.ffn_gate\.weight$=q3_k
+^blk\.2\.ffn_up\.weight$=q3_k
+^blk\.3\.attn_k\.weight$=q4_k
+^blk\.3\.attn_output\.weight$=q5_k
+^blk\.3\.attn_q\.weight$=q3_k
+^blk\.3\.attn_v\.weight$=q5_k
+^blk\.3\.ffn_down\.weight$=q3_k
+^blk\.3\.ffn_gate\.weight$=q3_k
+^blk\.3\.ffn_up\.weight$=q3_k
+^blk\.4\.attn_k\.weight$=q4_k
+^blk\.4\.attn_output\.weight$=q5_k
+^blk\.4\.attn_q\.weight$=q3_k
+^blk\.4\.attn_v\.weight$=q5_k
+^blk\.4\.ffn_down\.weight$=q3_k
+^blk\.4\.ffn_gate\.weight$=q3_k
+^blk\.4\.ffn_up\.weight$=q3_k
+^blk\.5\.attn_k\.weight$=q4_k
+^blk\.5\.attn_output\.weight$=q5_k
+^blk\.5\.attn_q\.weight$=q3_k
+^blk\.5\.attn_v\.weight$=q5_k
+^blk\.5\.ffn_down\.weight$=q3_k
+^blk\.5\.ffn_gate\.weight$=q3_k
+^blk\.5\.ffn_up\.weight$=q3_k
+^blk\.6\.attn_k\.weight$=q4_k
+^blk\.6\.attn_output\.weight$=q5_k
+^blk\.6\.attn_q\.weight$=q3_k
+^blk\.6\.attn_v\.weight$=q5_k
+^blk\.6\.ffn_down\.weight$=q3_k
+^blk\.6\.ffn_gate\.weight$=q3_k
+^blk\.6\.ffn_up\.weight$=q3_k
+^blk\.7\.attn_k\.weight$=q4_k
+^blk\.7\.attn_output\.weight$=q5_k
+^blk\.7\.attn_q\.weight$=q3_k
+^blk\.7\.attn_v\.weight$=q5_k
+^blk\.7\.ffn_down\.weight$=q3_k
+^blk\.7\.ffn_gate\.weight$=q3_k
+^blk\.7\.ffn_up\.weight$=q3_k
+^blk\.8\.attn_k\.weight$=q4_k
+^blk\.8\.attn_output\.weight$=q5_k
+^blk\.8\.attn_q\.weight$=q3_k
+^blk\.8\.attn_v\.weight$=q5_k
+^blk\.8\.ffn_down\.weight$=q3_k
+^blk\.8\.ffn_gate\.weight$=q3_k
+^blk\.8\.ffn_up\.weight$=q3_k
+^blk\.9\.attn_k\.weight$=q5_k
+^blk\.9\.attn_output\.weight$=q4_k
+^blk\.9\.attn_q\.weight$=q3_k
+^blk\.9\.attn_v\.weight$=q5_k
+^blk\.9\.ffn_down\.weight$=q3_k
+^blk\.9\.ffn_gate\.weight$=q3_k
+^blk\.9\.ffn_up\.weight$=q3_k
+^blk\.10\.attn_k\.weight$=q5_k
+^blk\.10\.attn_output\.weight$=q4_k
+^blk\.10\.attn_q\.weight$=q3_k
+^blk\.10\.attn_v\.weight$=q5_k
+^blk\.10\.ffn_down\.weight$=q3_k
+^blk\.10\.ffn_gate\.weight$=q3_k
+^blk\.10\.ffn_up\.weight$=q3_k
+^blk\.11\.attn_k\.weight$=q4_k
+^blk\.11\.attn_output\.weight$=q3_k
+^blk\.11\.attn_q\.weight$=q3_k
+^blk\.11\.attn_v\.weight$=q5_k
+^blk\.11\.ffn_down\.weight$=q3_k
+^blk\.11\.ffn_gate\.weight$=q3_k
+^blk\.11\.ffn_up\.weight$=q3_k
+^blk\.12\.attn_k\.weight$=q4_k
+^blk\.12\.attn_output\.weight$=q3_k
+^blk\.12\.attn_q\.weight$=q3_k
+^blk\.12\.attn_v\.weight$=q5_k
+^blk\.12\.ffn_down\.weight$=q3_k
+^blk\.12\.ffn_gate\.weight$=q3_k
+^blk\.12\.ffn_up\.weight$=q3_k
+^blk\.13\.attn_k\.weight$=q4_k
+^blk\.13\.attn_output\.weight$=q3_k
+^blk\.13\.attn_q\.weight$=q3_k
+^blk\.13\.attn_v\.weight$=q5_k
+^blk\.13\.ffn_down\.weight$=q3_k
+^blk\.13\.ffn_gate\.weight$=q3_k
+^blk\.13\.ffn_up\.weight$=q3_k
+^blk\.14\.attn_k\.weight$=q4_k
+^blk\.14\.attn_output\.weight$=q3_k
+^blk\.14\.attn_q\.weight$=q3_k
+^blk\.14\.attn_v\.weight$=q5_k
+^blk\.14\.ffn_down\.weight$=q3_k
+^blk\.14\.ffn_gate\.weight$=q3_k
+^blk\.14\.ffn_up\.weight$=q3_k
+^blk\.15\.attn_k\.weight$=q4_k
+^blk\.15\.attn_output\.weight$=q3_k
+^blk\.15\.attn_q\.weight$=q3_k
+^blk\.15\.attn_v\.weight$=q5_k
+^blk\.15\.ffn_down\.weight$=q3_k
+^blk\.15\.ffn_gate\.weight$=q3_k
+^blk\.15\.ffn_up\.weight$=q3_k
+^blk\.16\.attn_k\.weight$=q4_k
+^blk\.16\.attn_output\.weight$=q3_k
+^blk\.16\.attn_q\.weight$=q3_k
+^blk\.16\.attn_v\.weight$=q5_k
+^blk\.16\.ffn_down\.weight$=q3_k
+^blk\.16\.ffn_gate\.weight$=q3_k
+^blk\.16\.ffn_up\.weight$=q3_k
+^blk\.17\.attn_k\.weight$=q4_k
+^blk\.17\.attn_output\.weight$=q3_k
+^blk\.17\.attn_q\.weight$=q3_k
+^blk\.17\.attn_v\.weight$=q5_k
+^blk\.17\.ffn_down\.weight$=q3_k
+^blk\.17\.ffn_gate\.weight$=q3_k
+^blk\.17\.ffn_up\.weight$=q3_k
+^blk\.18\.attn_k\.weight$=q4_k
+^blk\.18\.attn_output\.weight$=q3_k
+^blk\.18\.attn_q\.weight$=q3_k
+^blk\.18\.attn_v\.weight$=q5_k
+^blk\.18\.ffn_down\.weight$=q3_k
+^blk\.18\.ffn_gate\.weight$=q3_k
+^blk\.18\.ffn_up\.weight$=q3_k
+^blk\.19\.attn_k\.weight$=q4_k
+^blk\.19\.attn_output\.weight$=q3_k
+^blk\.19\.attn_q\.weight$=q3_k
+^blk\.19\.attn_v\.weight$=q5_k
+^blk\.19\.ffn_down\.weight$=q3_k
+^blk\.19\.ffn_gate\.weight$=q3_k
+^blk\.19\.ffn_up\.weight$=q3_k
+^blk\.20\.attn_k\.weight$=q4_k
+^blk\.20\.attn_output\.weight$=q3_k
+^blk\.20\.attn_q\.weight$=q3_k
+^blk\.20\.attn_v\.weight$=q5_k
+^blk\.20\.ffn_down\.weight$=q3_k
+^blk\.20\.ffn_gate\.weight$=q3_k
+^blk\.20\.ffn_up\.weight$=q3_k
+^blk\.21\.attn_k\.weight$=q4_k
+^blk\.21\.attn_output\.weight$=q3_k
+^blk\.21\.attn_q\.weight$=q3_k
+^blk\.21\.attn_v\.weight$=q5_k
+^blk\.21\.ffn_down\.weight$=q3_k
+^blk\.21\.ffn_gate\.weight$=q3_k
+^blk\.21\.ffn_up\.weight$=q3_k
+^blk\.22\.attn_k\.weight$=q4_k
+^blk\.22\.attn_output\.weight$=q3_k
+^blk\.22\.attn_q\.weight$=q3_k
+^blk\.22\.attn_v\.weight$=q5_k
+^blk\.22\.ffn_down\.weight$=q3_k
+^blk\.22\.ffn_gate\.weight$=q3_k
+^blk\.22\.ffn_up\.weight$=q3_k
+^blk\.23\.attn_k\.weight$=q4_k
+^blk\.23\.attn_output\.weight$=q3_k
+^blk\.23\.attn_q\.weight$=q3_k
+^blk\.23\.attn_v\.weight$=q5_k
+^blk\.23\.ffn_down\.weight$=q3_k
+^blk\.23\.ffn_gate\.weight$=q3_k
+^blk\.23\.ffn_up\.weight$=q3_k
+^blk\.24\.attn_k\.weight$=q4_k
+^blk\.24\.attn_output\.weight$=q3_k
+^blk\.24\.attn_q\.weight$=q3_k
+^blk\.24\.attn_v\.weight$=q5_k
+^blk\.24\.ffn_down\.weight$=q3_k
+^blk\.24\.ffn_gate\.weight$=q3_k
+^blk\.24\.ffn_up\.weight$=q3_k
+^blk\.25\.attn_k\.weight$=q4_k
+^blk\.25\.attn_output\.weight$=q3_k
+^blk\.25\.attn_q\.weight$=q3_k
+^blk\.25\.attn_v\.weight$=q5_k
+^blk\.25\.ffn_down\.weight$=q3_k
+^blk\.25\.ffn_gate\.weight$=q3_k
+^blk\.25\.ffn_up\.weight$=q3_k
+^blk\.26\.attn_k\.weight$=q4_k
+^blk\.26\.attn_output\.weight$=q3_k
+^blk\.26\.attn_q\.weight$=q3_k
+^blk\.26\.attn_v\.weight$=q5_k
+^blk\.26\.ffn_down\.weight$=q3_k
+^blk\.26\.ffn_gate\.weight$=q3_k
+^blk\.26\.ffn_up\.weight$=q3_k
+^blk\.27\.attn_k\.weight$=q3_k
+^blk\.27\.attn_output\.weight$=q3_k
+^blk\.27\.attn_q\.weight$=q3_k
+^blk\.27\.attn_v\.weight$=q5_k
+^blk\.27\.ffn_down\.weight$=q3_k
+^blk\.27\.ffn_gate\.weight$=q3_k
+^blk\.27\.ffn_up\.weight$=q3_k
+^blk\.28\.attn_k\.weight$=q3_k
+^blk\.28\.attn_output\.weight$=q3_k
+^blk\.28\.attn_q\.weight$=q3_k
+^blk\.28\.attn_v\.weight$=q5_k
+^blk\.28\.ffn_down\.weight$=q3_k
+^blk\.28\.ffn_gate\.weight$=q3_k
+^blk\.28\.ffn_up\.weight$=q3_k
+^blk\.29\.attn_k\.weight$=q3_k
+^blk\.29\.attn_output\.weight$=q3_k
+^blk\.29\.attn_q\.weight$=q3_k
+^blk\.29\.attn_v\.weight$=q5_k
+^blk\.29\.ffn_down\.weight$=q3_k
+^blk\.29\.ffn_gate\.weight$=q3_k
+^blk\.29\.ffn_up\.weight$=q3_k
+^blk\.30\.attn_k\.weight$=q3_k
+^blk\.30\.attn_output\.weight$=q3_k
+^blk\.30\.attn_q\.weight$=q3_k
+^blk\.30\.attn_v\.weight$=q5_k
+^blk\.30\.ffn_down\.weight$=q3_k
+^blk\.30\.ffn_gate\.weight$=q3_k
+^blk\.30\.ffn_up\.weight$=q3_k
+^blk\.31\.attn_k\.weight$=q3_k
+^blk\.31\.attn_output\.weight$=q3_k
+^blk\.31\.attn_q\.weight$=q3_k
+^blk\.31\.attn_v\.weight$=q5_k
+^blk\.31\.ffn_down\.weight$=q3_k
+^blk\.31\.ffn_gate\.weight$=q3_k
+^blk\.31\.ffn_up\.weight$=q3_k
+^blk\.32\.attn_k\.weight$=q3_k
+^blk\.32\.attn_output\.weight$=q3_k
+^blk\.32\.attn_q\.weight$=q3_k
+^blk\.32\.attn_v\.weight$=q5_k
+^blk\.32\.ffn_down\.weight$=q3_k
+^blk\.32\.ffn_gate\.weight$=q3_k
+^blk\.32\.ffn_up\.weight$=q3_k
+^blk\.33\.attn_k\.weight$=q3_k
+^blk\.33\.attn_output\.weight$=q3_k
+^blk\.33\.attn_q\.weight$=q3_k
+^blk\.33\.attn_v\.weight$=q5_k
+^blk\.33\.ffn_down\.weight$=q3_k
+^blk\.33\.ffn_gate\.weight$=q3_k
+^blk\.33\.ffn_up\.weight$=q3_k
+^blk\.34\.attn_k\.weight$=q3_k
+^blk\.34\.attn_output\.weight$=q3_k
+^blk\.34\.attn_q\.weight$=q3_k
+^blk\.34\.attn_v\.weight$=q5_k
+^blk\.34\.ffn_down\.weight$=q3_k
+^blk\.34\.ffn_gate\.weight$=q3_k
+^blk\.34\.ffn_up\.weight$=q3_k
+^blk\.35\.attn_k\.weight$=q3_k
+^blk\.35\.attn_output\.weight$=q3_k
+^blk\.35\.attn_q\.weight$=q3_k
+^blk\.35\.attn_v\.weight$=q5_k
+^blk\.35\.ffn_down\.weight$=q3_k
+^blk\.35\.ffn_gate\.weight$=q3_k
+^blk\.35\.ffn_up\.weight$=q4_k

layouts/AesCode-8B-Q4_K_L.layout.json

@@ -0,0 +1,89 @@
+{
+  "model": "/media/m2teamgroup/models/AesCode-8B-GGUF/AesCode-8B-bf16.gguf",
+  "imatrix": "/media/m2teamgroup/models/AesCode-8B-GGUF/AesCode-8B-imatrix.gguf",
+  "ftype": "Q4_K_L",
+  "base_type": "q4_k",
+  "family": "k",
+  "recipe": "ident",
+  "variant": "shr0.50",
+  "auto_crush": false,
+  "rung_mode": "share",
+  "tier": "L",
+  "rung_share": 0.5,
+  "crush_pass_bytes": null,
+  "embd_shares": {
+    "output.weight": {
+      "share": 0.0983,
+      "type": "q6_k"
+    },
+    "token_embd.weight": {
+      "share": 0.0674,
+      "type": "q4_k"
+    }
+  },
+  "arch": "dense",
+  "tied": false,
+  "n_layer": 36,
+  "body_rule": true,
+  "pins": {},
+  "shape_pins": {},
+  "fixed_pins": {},
+  "embeddings": {
+    "output.weight": "q6_k",
+    "token_embd.weight": "q4_k"
+  },
+  "bytes": {
+    "heuristic_body+embd": 5192822784,
+    "map_body+embd": 5578928128,
+    "ratio": 1.0744,
+    "note": "quantizable, unpinned tensors only; pinned/unquantized tensors are identical in both"
+  },
+  "body_base_share": 0.5,
+  "body_bpw": 5.435,
+  "body_bpw_heuristic": 4.99,
+  "total_params": 8190735360,
+  "file_bytes_pred": 5580161024,
+  "file_bpw_pred": 5.45,
+  "histogram": {
+    "attn_k": {
+      "q8_0": 27,
+      "q6_k": 9
+    },
+    "attn_output": {
+      "q6_k": 20,
+      "q8_0": 8,
+      "q4_k": 8
+    },
+    "attn_q": {
+      "q4_k": 35,
+      "q6_k": 1
+    },
+    "attn_v": {
+      "q8_0": 36
+    },
+    "ffn_down": {
+      "q6_k": 18,
+      "q4_k": 18
+    },
+    "ffn_gate": {
+      "q4_k": 33,
+      "q6_k": 2,
+      "q5_k": 1
+    },
+    "ffn_up": {
+      "q4_k": 18,
+      "q6_k": 17,
+      "q8_0": 1
+    }
+  },
+  "prior": "layout/prior.json",
+  "pin_rule": "floor",
+  "embd_cap": "heur",
+  "generator": "auto_quant_v2.layout",
+  "generator_key": "3bf8b43e20a68f0b",
+  "quant_type": "Q4_K_L",
+  "quantize_ftype": "Q4_K_M",
+  "llama_cpp_version": "b11490",
+  "unknown_kinds": {},
+  "commit": "5d550ef5561a30307f1d436db2c6835b75395e77"
+}

layouts/AesCode-8B-Q4_K_L.tensor-types.txt

@@ -0,0 +1,254 @@
+^output\.weight$=q6_k
+^token_embd\.weight$=q4_k
+^blk\.0\.attn_k\.weight$=q8_0
+^blk\.0\.attn_output\.weight$=q6_k
+^blk\.0\.attn_q\.weight$=q4_k
+^blk\.0\.attn_v\.weight$=q8_0
+^blk\.0\.ffn_down\.weight$=q6_k
+^blk\.0\.ffn_gate\.weight$=q4_k
+^blk\.0\.ffn_up\.weight$=q6_k
+^blk\.1\.attn_k\.weight$=q8_0
+^blk\.1\.attn_output\.weight$=q8_0
+^blk\.1\.attn_q\.weight$=q4_k
+^blk\.1\.attn_v\.weight$=q8_0
+^blk\.1\.ffn_down\.weight$=q6_k
+^blk\.1\.ffn_gate\.weight$=q6_k
+^blk\.1\.ffn_up\.weight$=q6_k
+^blk\.2\.attn_k\.weight$=q8_0
+^blk\.2\.attn_output\.weight$=q8_0
+^blk\.2\.attn_q\.weight$=q4_k
+^blk\.2\.attn_v\.weight$=q8_0
+^blk\.2\.ffn_down\.weight$=q6_k
+^blk\.2\.ffn_gate\.weight$=q5_k
+^blk\.2\.ffn_up\.weight$=q6_k
+^blk\.3\.attn_k\.weight$=q8_0
+^blk\.3\.attn_output\.weight$=q8_0
+^blk\.3\.attn_q\.weight$=q4_k
+^blk\.3\.attn_v\.weight$=q8_0
+^blk\.3\.ffn_down\.weight$=q6_k
+^blk\.3\.ffn_gate\.weight$=q4_k
+^blk\.3\.ffn_up\.weight$=q6_k
+^blk\.4\.attn_k\.weight$=q8_0
+^blk\.4\.attn_output\.weight$=q8_0
+^blk\.4\.attn_q\.weight$=q4_k
+^blk\.4\.attn_v\.weight$=q8_0
+^blk\.4\.ffn_down\.weight$=q6_k
+^blk\.4\.ffn_gate\.weight$=q4_k
+^blk\.4\.ffn_up\.weight$=q6_k
+^blk\.5\.attn_k\.weight$=q8_0
+^blk\.5\.attn_output\.weight$=q8_0
+^blk\.5\.attn_q\.weight$=q4_k
+^blk\.5\.attn_v\.weight$=q8_0
+^blk\.5\.ffn_down\.weight$=q6_k
+^blk\.5\.ffn_gate\.weight$=q4_k
+^blk\.5\.ffn_up\.weight$=q6_k
+^blk\.6\.attn_k\.weight$=q8_0
+^blk\.6\.attn_output\.weight$=q8_0
+^blk\.6\.attn_q\.weight$=q4_k
+^blk\.6\.attn_v\.weight$=q8_0
+^blk\.6\.ffn_down\.weight$=q6_k
+^blk\.6\.ffn_gate\.weight$=q4_k
+^blk\.6\.ffn_up\.weight$=q6_k
+^blk\.7\.attn_k\.weight$=q8_0
+^blk\.7\.attn_output\.weight$=q8_0
+^blk\.7\.attn_q\.weight$=q4_k
+^blk\.7\.attn_v\.weight$=q8_0
+^blk\.7\.ffn_down\.weight$=q6_k
+^blk\.7\.ffn_gate\.weight$=q4_k
+^blk\.7\.ffn_up\.weight$=q6_k
+^blk\.8\.attn_k\.weight$=q8_0
+^blk\.8\.attn_output\.weight$=q8_0
+^blk\.8\.attn_q\.weight$=q4_k
+^blk\.8\.attn_v\.weight$=q8_0
+^blk\.8\.ffn_down\.weight$=q6_k
+^blk\.8\.ffn_gate\.weight$=q4_k
+^blk\.8\.ffn_up\.weight$=q6_k
+^blk\.9\.attn_k\.weight$=q8_0
+^blk\.9\.attn_output\.weight$=q6_k
+^blk\.9\.attn_q\.weight$=q6_k
+^blk\.9\.attn_v\.weight$=q8_0
+^blk\.9\.ffn_down\.weight$=q4_k
+^blk\.9\.ffn_gate\.weight$=q4_k
+^blk\.9\.ffn_up\.weight$=q4_k
+^blk\.10\.attn_k\.weight$=q8_0
+^blk\.10\.attn_output\.weight$=q6_k
+^blk\.10\.attn_q\.weight$=q4_k
+^blk\.10\.attn_v\.weight$=q8_0
+^blk\.10\.ffn_down\.weight$=q4_k
+^blk\.10\.ffn_gate\.weight$=q4_k
+^blk\.10\.ffn_up\.weight$=q4_k
+^blk\.11\.attn_k\.weight$=q8_0
+^blk\.11\.attn_output\.weight$=q6_k
+^blk\.11\.attn_q\.weight$=q4_k
+^blk\.11\.attn_v\.weight$=q8_0
+^blk\.11\.ffn_down\.weight$=q4_k
+^blk\.11\.ffn_gate\.weight$=q4_k
+^blk\.11\.ffn_up\.weight$=q4_k
+^blk\.12\.attn_k\.weight$=q8_0
+^blk\.12\.attn_output\.weight$=q6_k
+^blk\.12\.attn_q\.weight$=q4_k
+^blk\.12\.attn_v\.weight$=q8_0
+^blk\.12\.ffn_down\.weight$=q4_k
+^blk\.12\.ffn_gate\.weight$=q4_k
+^blk\.12\.ffn_up\.weight$=q4_k
+^blk\.13\.attn_k\.weight$=q8_0
+^blk\.13\.attn_output\.weight$=q6_k
+^blk\.13\.attn_q\.weight$=q4_k
+^blk\.13\.attn_v\.weight$=q8_0
+^blk\.13\.ffn_down\.weight$=q4_k
+^blk\.13\.ffn_gate\.weight$=q4_k
+^blk\.13\.ffn_up\.weight$=q4_k
+^blk\.14\.attn_k\.weight$=q8_0
+^blk\.14\.attn_output\.weight$=q6_k
+^blk\.14\.attn_q\.weight$=q4_k
+^blk\.14\.attn_v\.weight$=q8_0
+^blk\.14\.ffn_down\.weight$=q4_k
+^blk\.14\.ffn_gate\.weight$=q4_k
+^blk\.14\.ffn_up\.weight$=q4_k
+^blk\.15\.attn_k\.weight$=q8_0
+^blk\.15\.attn_output\.weight$=q6_k
+^blk\.15\.attn_q\.weight$=q4_k
+^blk\.15\.attn_v\.weight$=q8_0
+^blk\.15\.ffn_down\.weight$=q4_k
+^blk\.15\.ffn_gate\.weight$=q4_k
+^blk\.15\.ffn_up\.weight$=q4_k
+^blk\.16\.attn_k\.weight$=q8_0
+^blk\.16\.attn_output\.weight$=q6_k
+^blk\.16\.attn_q\.weight$=q4_k
+^blk\.16\.attn_v\.weight$=q8_0
+^blk\.16\.ffn_down\.weight$=q4_k
+^blk\.16\.ffn_gate\.weight$=q4_k
+^blk\.16\.ffn_up\.weight$=q4_k
+^blk\.17\.attn_k\.weight$=q8_0
+^blk\.17\.attn_output\.weight$=q6_k
+^blk\.17\.attn_q\.weight$=q4_k
+^blk\.17\.attn_v\.weight$=q8_0
+^blk\.17\.ffn_down\.weight$=q4_k
+^blk\.17\.ffn_gate\.weight$=q4_k
+^blk\.17\.ffn_up\.weight$=q4_k
+^blk\.18\.attn_k\.weight$=q8_0
+^blk\.18\.attn_output\.weight$=q6_k
+^blk\.18\.attn_q\.weight$=q4_k
+^blk\.18\.attn_v\.weight$=q8_0
+^blk\.18\.ffn_down\.weight$=q4_k
+^blk\.18\.ffn_gate\.weight$=q4_k
+^blk\.18\.ffn_up\.weight$=q4_k
+^blk\.19\.attn_k\.weight$=q8_0
+^blk\.19\.attn_output\.weight$=q6_k
+^blk\.19\.attn_q\.weight$=q4_k
+^blk\.19\.attn_v\.weight$=q8_0
+^blk\.19\.ffn_down\.weight$=q4_k
+^blk\.19\.ffn_gate\.weight$=q4_k
+^blk\.19\.ffn_up\.weight$=q4_k
+^blk\.20\.attn_k\.weight$=q8_0
+^blk\.20\.attn_output\.weight$=q6_k
+^blk\.20\.attn_q\.weight$=q4_k
+^blk\.20\.attn_v\.weight$=q8_0
+^blk\.20\.ffn_down\.weight$=q4_k
+^blk\.20\.ffn_gate\.weight$=q4_k
+^blk\.20\.ffn_up\.weight$=q4_k
+^blk\.21\.attn_k\.weight$=q8_0
+^blk\.21\.attn_output\.weight$=q6_k
+^blk\.21\.attn_q\.weight$=q4_k
+^blk\.21\.attn_v\.weight$=q8_0
+^blk\.21\.ffn_down\.weight$=q4_k
+^blk\.21\.ffn_gate\.weight$=q4_k
+^blk\.21\.ffn_up\.weight$=q4_k
+^blk\.22\.attn_k\.weight$=q8_0
+^blk\.22\.attn_output\.weight$=q6_k
+^blk\.22\.attn_q\.weight$=q4_k
+^blk\.22\.attn_v\.weight$=q8_0
+^blk\.22\.ffn_down\.weight$=q4_k
+^blk\.22\.ffn_gate\.weight$=q4_k
+^blk\.22\.ffn_up\.weight$=q4_k
+^blk\.23\.attn_k\.weight$=q8_0
+^blk\.23\.attn_output\.weight$=q6_k
+^blk\.23\.attn_q\.weight$=q4_k
+^blk\.23\.attn_v\.weight$=q8_0
+^blk\.23\.ffn_down\.weight$=q4_k
+^blk\.23\.ffn_gate\.weight$=q4_k
+^blk\.23\.ffn_up\.weight$=q4_k
+^blk\.24\.attn_k\.weight$=q8_0
+^blk\.24\.attn_output\.weight$=q6_k
+^blk\.24\.attn_q\.weight$=q4_k
+^blk\.24\.attn_v\.weight$=q8_0
+^blk\.24\.ffn_down\.weight$=q4_k
+^blk\.24\.ffn_gate\.weight$=q4_k
+^blk\.24\.ffn_up\.weight$=q4_k
+^blk\.25\.attn_k\.weight$=q8_0
+^blk\.25\.attn_output\.weight$=q6_k
+^blk\.25\.attn_q\.weight$=q4_k
+^blk\.25\.attn_v\.weight$=q8_0
+^blk\.25\.ffn_down\.weight$=q4_k
+^blk\.25\.ffn_gate\.weight$=q4_k
+^blk\.25\.ffn_up\.weight$=q4_k
+^blk\.26\.attn_k\.weight$=q8_0
+^blk\.26\.attn_output\.weight$=q6_k
+^blk\.26\.attn_q\.weight$=q4_k
+^blk\.26\.attn_v\.weight$=q8_0
+^blk\.26\.ffn_down\.weight$=q4_k
+^blk\.26\.ffn_gate\.weight$=q4_k
+^blk\.26\.ffn_up\.weight$=q4_k
+^blk\.27\.attn_k\.weight$=q6_k
+^blk\.27\.attn_output\.weight$=q4_k
+^blk\.27\.attn_q\.weight$=q4_k
+^blk\.27\.attn_v\.weight$=q8_0
+^blk\.27\.ffn_down\.weight$=q6_k
+^blk\.27\.ffn_gate\.weight$=q4_k
+^blk\.27\.ffn_up\.weight$=q6_k
+^blk\.28\.attn_k\.weight$=q6_k
+^blk\.28\.attn_output\.weight$=q4_k
+^blk\.28\.attn_q\.weight$=q4_k
+^blk\.28\.attn_v\.weight$=q8_0
+^blk\.28\.ffn_down\.weight$=q6_k
+^blk\.28\.ffn_gate\.weight$=q4_k
+^blk\.28\.ffn_up\.weight$=q6_k
+^blk\.29\.attn_k\.weight$=q6_k
+^blk\.29\.attn_output\.weight$=q4_k
+^blk\.29\.attn_q\.weight$=q4_k
+^blk\.29\.attn_v\.weight$=q8_0
+^blk\.29\.ffn_down\.weight$=q6_k
+^blk\.29\.ffn_gate\.weight$=q4_k
+^blk\.29\.ffn_up\.weight$=q6_k
+^blk\.30\.attn_k\.weight$=q6_k
+^blk\.30\.attn_output\.weight$=q4_k
+^blk\.30\.attn_q\.weight$=q4_k
+^blk\.30\.attn_v\.weight$=q8_0
+^blk\.30\.ffn_down\.weight$=q6_k
+^blk\.30\.ffn_gate\.weight$=q4_k
+^blk\.30\.ffn_up\.weight$=q6_k
+^blk\.31\.attn_k\.weight$=q6_k
+^blk\.31\.attn_output\.weight$=q4_k
+^blk\.31\.attn_q\.weight$=q4_k
+^blk\.31\.attn_v\.weight$=q8_0
+^blk\.31\.ffn_down\.weight$=q6_k
+^blk\.31\.ffn_gate\.weight$=q4_k
+^blk\.31\.ffn_up\.weight$=q6_k
+^blk\.32\.attn_k\.weight$=q6_k
+^blk\.32\.attn_output\.weight$=q4_k
+^blk\.32\.attn_q\.weight$=q4_k
+^blk\.32\.attn_v\.weight$=q8_0
+^blk\.32\.ffn_down\.weight$=q6_k
+^blk\.32\.ffn_gate\.weight$=q4_k
+^blk\.32\.ffn_up\.weight$=q6_k
+^blk\.33\.attn_k\.weight$=q6_k
+^blk\.33\.attn_output\.weight$=q4_k
+^blk\.33\.attn_q\.weight$=q4_k
+^blk\.33\.attn_v\.weight$=q8_0
+^blk\.33\.ffn_down\.weight$=q6_k
+^blk\.33\.ffn_gate\.weight$=q4_k
+^blk\.33\.ffn_up\.weight$=q6_k
+^blk\.34\.attn_k\.weight$=q6_k
+^blk\.34\.attn_output\.weight$=q4_k
+^blk\.34\.attn_q\.weight$=q4_k
+^blk\.34\.attn_v\.weight$=q8_0
+^blk\.34\.ffn_down\.weight$=q6_k
+^blk\.34\.ffn_gate\.weight$=q4_k
+^blk\.34\.ffn_up\.weight$=q6_k
+^blk\.35\.attn_k\.weight$=q6_k
+^blk\.35\.attn_output\.weight$=q6_k
+^blk\.35\.attn_q\.weight$=q4_k
+^blk\.35\.attn_v\.weight$=q8_0
+^blk\.35\.ffn_down\.weight$=q6_k
+^blk\.35\.ffn_gate\.weight$=q6_k
+^blk\.35\.ffn_up\.weight$=q8_0

layouts/AesCode-8B-Q4_K_M.layout.json

@@ -0,0 +1,86 @@
+{
+  "model": "/media/m2teamgroup/models/AesCode-8B-GGUF/AesCode-8B-bf16.gguf",
+  "imatrix": "/media/m2teamgroup/models/AesCode-8B-GGUF/AesCode-8B-imatrix.gguf",
+  "ftype": "Q4_K_M",
+  "base_type": "q4_k",
+  "family": "k",
+  "recipe": "ident",
+  "variant": "shr0.70",
+  "auto_crush": false,
+  "rung_mode": "share",
+  "tier": "M",
+  "rung_share": 0.7,
+  "crush_pass_bytes": null,
+  "embd_shares": {
+    "output.weight": {
+      "share": 0.0983,
+      "type": "q6_k"
+    },
+    "token_embd.weight": {
+      "share": 0.0674,
+      "type": "q4_k"
+    }
+  },
+  "arch": "dense",
+  "tied": false,
+  "n_layer": 36,
+  "body_rule": true,
+  "pins": {},
+  "shape_pins": {},
+  "fixed_pins": {},
+  "embeddings": {
+    "output.weight": "q6_k",
+    "token_embd.weight": "q4_k"
+  },
+  "bytes": {
+    "heuristic_body+embd": 5192822784,
+    "map_body+embd": 5204324352,
+    "ratio": 1.0022,
+    "note": "quantizable, unpinned tensors only; pinned/unquantized tensors are identical in both"
+  },
+  "body_base_share": 0.7,
+  "body_bpw": 5.003,
+  "body_bpw_heuristic": 4.99,
+  "total_params": 8190735360,
+  "file_bytes_pred": 5205557248,
+  "file_bpw_pred": 5.084,
+  "histogram": {
+    "attn_k": {
+      "q6_k": 25,
+      "q4_k": 9,
+      "q8_0": 2
+    },
+    "attn_output": {
+      "q4_k": 14,
+      "q5_k": 14,
+      "q8_0": 8
+    },
+    "attn_q": {
+      "q4_k": 36
+    },
+    "attn_v": {
+      "q8_0": 36
+    },
+    "ffn_down": {
+      "q4_k": 27,
+      "q6_k": 9
+    },
+    "ffn_gate": {
+      "q4_k": 36
+    },
+    "ffn_up": {
+      "q4_k": 27,
+      "q6_k": 9
+    }
+  },
+  "prior": "layout/prior.json",
+  "pin_rule": "floor",
+  "embd_cap": "heur",
+  "generator": "auto_quant_v2.layout",
+  "generator_key": "3bf8b43e20a68f0b",
+  "quant_type": "Q4_K_M",
+  "quantize_ftype": "Q4_K_M",
+  "llama_cpp_version": "b11490",
+  "unknown_kinds": {},
+  "commit": "5d550ef5561a30307f1d436db2c6835b75395e77"
+}

layouts/AesCode-8B-Q4_K_M.tensor-types.txt

@@ -0,0 +1,254 @@
+^output\.weight$=q6_k
+^token_embd\.weight$=q4_k
+^blk\.0\.attn_k\.weight$=q8_0
+^blk\.0\.attn_output\.weight$=q4_k
+^blk\.0\.attn_q\.weight$=q4_k
+^blk\.0\.attn_v\.weight$=q8_0
+^blk\.0\.ffn_down\.weight$=q4_k
+^blk\.0\.ffn_gate\.weight$=q4_k
+^blk\.0\.ffn_up\.weight$=q4_k
+^blk\.1\.attn_k\.weight$=q6_k
+^blk\.1\.attn_output\.weight$=q8_0
+^blk\.1\.attn_q\.weight$=q4_k
+^blk\.1\.attn_v\.weight$=q8_0
+^blk\.1\.ffn_down\.weight$=q6_k
+^blk\.1\.ffn_gate\.weight$=q4_k
+^blk\.1\.ffn_up\.weight$=q6_k
+^blk\.2\.attn_k\.weight$=q6_k
+^blk\.2\.attn_output\.weight$=q8_0
+^blk\.2\.attn_q\.weight$=q4_k
+^blk\.2\.attn_v\.weight$=q8_0
+^blk\.2\.ffn_down\.weight$=q6_k
+^blk\.2\.ffn_gate\.weight$=q4_k
+^blk\.2\.ffn_up\.weight$=q6_k
+^blk\.3\.attn_k\.weight$=q6_k
+^blk\.3\.attn_output\.weight$=q8_0
+^blk\.3\.attn_q\.weight$=q4_k
+^blk\.3\.attn_v\.weight$=q8_0
+^blk\.3\.ffn_down\.weight$=q6_k
+^blk\.3\.ffn_gate\.weight$=q4_k
+^blk\.3\.ffn_up\.weight$=q6_k
+^blk\.4\.attn_k\.weight$=q6_k
+^blk\.4\.attn_output\.weight$=q8_0
+^blk\.4\.attn_q\.weight$=q4_k
+^blk\.4\.attn_v\.weight$=q8_0
+^blk\.4\.ffn_down\.weight$=q6_k
+^blk\.4\.ffn_gate\.weight$=q4_k
+^blk\.4\.ffn_up\.weight$=q6_k
+^blk\.5\.attn_k\.weight$=q6_k
+^blk\.5\.attn_output\.weight$=q8_0
+^blk\.5\.attn_q\.weight$=q4_k
+^blk\.5\.attn_v\.weight$=q8_0
+^blk\.5\.ffn_down\.weight$=q6_k
+^blk\.5\.ffn_gate\.weight$=q4_k
+^blk\.5\.ffn_up\.weight$=q6_k
+^blk\.6\.attn_k\.weight$=q6_k
+^blk\.6\.attn_output\.weight$=q8_0
+^blk\.6\.attn_q\.weight$=q4_k
+^blk\.6\.attn_v\.weight$=q8_0
+^blk\.6\.ffn_down\.weight$=q6_k
+^blk\.6\.ffn_gate\.weight$=q4_k
+^blk\.6\.ffn_up\.weight$=q6_k
+^blk\.7\.attn_k\.weight$=q6_k
+^blk\.7\.attn_output\.weight$=q8_0
+^blk\.7\.attn_q\.weight$=q4_k
+^blk\.7\.attn_v\.weight$=q8_0
+^blk\.7\.ffn_down\.weight$=q6_k
+^blk\.7\.ffn_gate\.weight$=q4_k
+^blk\.7\.ffn_up\.weight$=q6_k
+^blk\.8\.attn_k\.weight$=q6_k
+^blk\.8\.attn_output\.weight$=q8_0
+^blk\.8\.attn_q\.weight$=q4_k
+^blk\.8\.attn_v\.weight$=q8_0
+^blk\.8\.ffn_down\.weight$=q6_k
+^blk\.8\.ffn_gate\.weight$=q4_k
+^blk\.8\.ffn_up\.weight$=q6_k
+^blk\.9\.attn_k\.weight$=q8_0
+^blk\.9\.attn_output\.weight$=q5_k
+^blk\.9\.attn_q\.weight$=q4_k
+^blk\.9\.attn_v\.weight$=q8_0
+^blk\.9\.ffn_down\.weight$=q4_k
+^blk\.9\.ffn_gate\.weight$=q4_k
+^blk\.9\.ffn_up\.weight$=q4_k
+^blk\.10\.attn_k\.weight$=q6_k
+^blk\.10\.attn_output\.weight$=q5_k
+^blk\.10\.attn_q\.weight$=q4_k
+^blk\.10\.attn_v\.weight$=q8_0
+^blk\.10\.ffn_down\.weight$=q4_k
+^blk\.10\.ffn_gate\.weight$=q4_k
+^blk\.10\.ffn_up\.weight$=q4_k
+^blk\.11\.attn_k\.weight$=q6_k
+^blk\.11\.attn_output\.weight$=q5_k
+^blk\.11\.attn_q\.weight$=q4_k
+^blk\.11\.attn_v\.weight$=q8_0
+^blk\.11\.ffn_down\.weight$=q4_k
+^blk\.11\.ffn_gate\.weight$=q4_k
+^blk\.11\.ffn_up\.weight$=q4_k
+^blk\.12\.attn_k\.weight$=q6_k
+^blk\.12\.attn_output\.weight$=q5_k
+^blk\.12\.attn_q\.weight$=q4_k
+^blk\.12\.attn_v\.weight$=q8_0
+^blk\.12\.ffn_down\.weight$=q4_k
+^blk\.12\.ffn_gate\.weight$=q4_k
+^blk\.12\.ffn_up\.weight$=q4_k
+^blk\.13\.attn_k\.weight$=q6_k
+^blk\.13\.attn_output\.weight$=q5_k
+^blk\.13\.attn_q\.weight$=q4_k
+^blk\.13\.attn_v\.weight$=q8_0
+^blk\.13\.ffn_down\.weight$=q4_k
+^blk\.13\.ffn_gate\.weight$=q4_k
+^blk\.13\.ffn_up\.weight$=q4_k
+^blk\.14\.attn_k\.weight$=q6_k
+^blk\.14\.attn_output\.weight$=q5_k
+^blk\.14\.attn_q\.weight$=q4_k
+^blk\.14\.attn_v\.weight$=q8_0
+^blk\.14\.ffn_down\.weight$=q4_k
+^blk\.14\.ffn_gate\.weight$=q4_k
+^blk\.14\.ffn_up\.weight$=q4_k
+^blk\.15\.attn_k\.weight$=q6_k
+^blk\.15\.attn_output\.weight$=q5_k
+^blk\.15\.attn_q\.weight$=q4_k
+^blk\.15\.attn_v\.weight$=q8_0
+^blk\.15\.ffn_down\.weight$=q4_k
+^blk\.15\.ffn_gate\.weight$=q4_k
+^blk\.15\.ffn_up\.weight$=q4_k
+^blk\.16\.attn_k\.weight$=q6_k
+^blk\.16\.attn_output\.weight$=q5_k
+^blk\.16\.attn_q\.weight$=q4_k
+^blk\.16\.attn_v\.weight$=q8_0
+^blk\.16\.ffn_down\.weight$=q4_k
+^blk\.16\.ffn_gate\.weight$=q4_k
+^blk\.16\.ffn_up\.weight$=q4_k
+^blk\.17\.attn_k\.weight$=q6_k
+^blk\.17\.attn_output\.weight$=q5_k
+^blk\.17\.attn_q\.weight$=q4_k
+^blk\.17\.attn_v\.weight$=q8_0
+^blk\.17\.ffn_down\.weight$=q4_k
+^blk\.17\.ffn_gate\.weight$=q4_k
+^blk\.17\.ffn_up\.weight$=q4_k
+^blk\.18\.attn_k\.weight$=q6_k
+^blk\.18\.attn_output\.weight$=q5_k
+^blk\.18\.attn_q\.weight$=q4_k
+^blk\.18\.attn_v\.weight$=q8_0
+^blk\.18\.ffn_down\.weight$=q4_k
+^blk\.18\.ffn_gate\.weight$=q4_k
+^blk\.18\.ffn_up\.weight$=q4_k
+^blk\.19\.attn_k\.weight$=q6_k
+^blk\.19\.attn_output\.weight$=q5_k
+^blk\.19\.attn_q\.weight$=q4_k
+^blk\.19\.attn_v\.weight$=q8_0
+^blk\.19\.ffn_down\.weight$=q4_k
+^blk\.19\.ffn_gate\.weight$=q4_k
+^blk\.19\.ffn_up\.weight$=q4_k
+^blk\.20\.attn_k\.weight$=q6_k
+^blk\.20\.attn_output\.weight$=q5_k
+^blk\.20\.attn_q\.weight$=q4_k
+^blk\.20\.attn_v\.weight$=q8_0
+^blk\.20\.ffn_down\.weight$=q4_k
+^blk\.20\.ffn_gate\.weight$=q4_k
+^blk\.20\.ffn_up\.weight$=q4_k
+^blk\.21\.attn_k\.weight$=q6_k
+^blk\.21\.attn_output\.weight$=q5_k
+^blk\.21\.attn_q\.weight$=q4_k
+^blk\.21\.attn_v\.weight$=q8_0
+^blk\.21\.ffn_down\.weight$=q4_k
+^blk\.21\.ffn_gate\.weight$=q4_k
+^blk\.21\.ffn_up\.weight$=q4_k
+^blk\.22\.attn_k\.weight$=q6_k
+^blk\.22\.attn_output\.weight$=q5_k
+^blk\.22\.attn_q\.weight$=q4_k
+^blk\.22\.attn_v\.weight$=q8_0
+^blk\.22\.ffn_down\.weight$=q4_k
+^blk\.22\.ffn_gate\.weight$=q4_k
+^blk\.22\.ffn_up\.weight$=q4_k
+^blk\.23\.attn_k\.weight$=q6_k
+^blk\.23\.attn_output\.weight$=q4_k
+^blk\.23\.attn_q\.weight$=q4_k
+^blk\.23\.attn_v\.weight$=q8_0
+^blk\.23\.ffn_down\.weight$=q4_k
+^blk\.23\.ffn_gate\.weight$=q4_k
+^blk\.23\.ffn_up\.weight$=q4_k
+^blk\.24\.attn_k\.weight$=q6_k
+^blk\.24\.attn_output\.weight$=q4_k
+^blk\.24\.attn_q\.weight$=q4_k
+^blk\.24\.attn_v\.weight$=q8_0
+^blk\.24\.ffn_down\.weight$=q4_k
+^blk\.24\.ffn_gate\.weight$=q4_k
+^blk\.24\.ffn_up\.weight$=q4_k
+^blk\.25\.attn_k\.weight$=q6_k
+^blk\.25\.attn_output\.weight$=q4_k
+^blk\.25\.attn_q\.weight$=q4_k
+^blk\.25\.attn_v\.weight$=q8_0
+^blk\.25\.ffn_down\.weight$=q4_k
+^blk\.25\.ffn_gate\.weight$=q4_k
+^blk\.25\.ffn_up\.weight$=q4_k
+^blk\.26\.attn_k\.weight$=q6_k
+^blk\.26\.attn_output\.weight$=q4_k
+^blk\.26\.attn_q\.weight$=q4_k
+^blk\.26\.attn_v\.weight$=q8_0
+^blk\.26\.ffn_down\.weight$=q4_k
+^blk\.26\.ffn_gate\.weight$=q4_k
+^blk\.26\.ffn_up\.weight$=q4_k
+^blk\.27\.attn_k\.weight$=q4_k
+^blk\.27\.attn_output\.weight$=q4_k
+^blk\.27\.attn_q\.weight$=q4_k
+^blk\.27\.attn_v\.weight$=q8_0
+^blk\.27\.ffn_down\.weight$=q4_k
+^blk\.27\.ffn_gate\.weight$=q4_k
+^blk\.27\.ffn_up\.weight$=q4_k
+^blk\.28\.attn_k\.weight$=q4_k
+^blk\.28\.attn_output\.weight$=q4_k
+^blk\.28\.attn_q\.weight$=q4_k
+^blk\.28\.attn_v\.weight$=q8_0
+^blk\.28\.ffn_down\.weight$=q4_k
+^blk\.28\.ffn_gate\.weight$=q4_k
+^blk\.28\.ffn_up\.weight$=q4_k
+^blk\.29\.attn_k\.weight$=q4_k
+^blk\.29\.attn_output\.weight$=q4_k
+^blk\.29\.attn_q\.weight$=q4_k
+^blk\.29\.attn_v\.weight$=q8_0
+^blk\.29\.ffn_down\.weight$=q4_k
+^blk\.29\.ffn_gate\.weight$=q4_k
+^blk\.29\.ffn_up\.weight$=q4_k
+^blk\.30\.attn_k\.weight$=q4_k
+^blk\.30\.attn_output\.weight$=q4_k
+^blk\.30\.attn_q\.weight$=q4_k
+^blk\.30\.attn_v\.weight$=q8_0
+^blk\.30\.ffn_down\.weight$=q4_k
+^blk\.30\.ffn_gate\.weight$=q4_k
+^blk\.30\.ffn_up\.weight$=q4_k
+^blk\.31\.attn_k\.weight$=q4_k
+^blk\.31\.attn_output\.weight$=q4_k
+^blk\.31\.attn_q\.weight$=q4_k
+^blk\.31\.attn_v\.weight$=q8_0
+^blk\.31\.ffn_down\.weight$=q4_k
+^blk\.31\.ffn_gate\.weight$=q4_k
+^blk\.31\.ffn_up\.weight$=q4_k
+^blk\.32\.attn_k\.weight$=q4_k
+^blk\.32\.attn_output\.weight$=q4_k
+^blk\.32\.attn_q\.weight$=q4_k
+^blk\.32\.attn_v\.weight$=q8_0
+^blk\.32\.ffn_down\.weight$=q4_k
+^blk\.32\.ffn_gate\.weight$=q4_k
+^blk\.32\.ffn_up\.weight$=q4_k
+^blk\.33\.attn_k\.weight$=q4_k
+^blk\.33\.attn_output\.weight$=q4_k
+^blk\.33\.attn_q\.weight$=q4_k
+^blk\.33\.attn_v\.weight$=q8_0
+^blk\.33\.ffn_down\.weight$=q4_k
+^blk\.33\.ffn_gate\.weight$=q4_k
+^blk\.33\.ffn_up\.weight$=q4_k
+^blk\.34\.attn_k\.weight$=q4_k
+^blk\.34\.attn_output\.weight$=q4_k
+^blk\.34\.attn_q\.weight$=q4_k
+^blk\.34\.attn_v\.weight$=q8_0
+^blk\.34\.ffn_down\.weight$=q4_k
+^blk\.34\.ffn_gate\.weight$=q4_k
+^blk\.34\.ffn_up\.weight$=q4_k
+^blk\.35\.attn_k\.weight$=q4_k
+^blk\.35\.attn_output\.weight$=q4_k
+^blk\.35\.attn_q\.weight$=q4_k
+^blk\.35\.attn_v\.weight$=q8_0
+^blk\.35\.ffn_down\.weight$=q6_k
+^blk\.35\.ffn_gate\.weight$=q4_k
+^blk\.35\.ffn_up\.weight$=q6_k

layouts/AesCode-8B-Q4_K_S.layout.json

@@ -0,0 +1,84 @@
+{
+  "model": "/media/m2teamgroup/models/AesCode-8B-GGUF/AesCode-8B-bf16.gguf",
+  "imatrix": "/media/m2teamgroup/models/AesCode-8B-GGUF/AesCode-8B-imatrix.gguf",
+  "ftype": "Q4_K_S",
+  "base_type": "q4_k",
+  "family": "k",
+  "recipe": "ident",
+  "variant": "shr0.90",
+  "auto_crush": false,
+  "rung_mode": "share",
+  "tier": "S",
+  "rung_share": 0.9,
+  "crush_pass_bytes": null,
+  "embd_shares": {
+    "output.weight": {
+      "share": 0.1046,
+      "type": "q6_k"
+    },
+    "token_embd.weight": {
+      "share": 0.0717,
+      "type": "q4_k"
+    }
+  },
+  "arch": "dense",
+  "tied": false,
+  "n_layer": 36,
+  "body_rule": true,
+  "pins": {},
+  "shape_pins": {},
+  "fixed_pins": {},
+  "embeddings": {
+    "output.weight": "q6_k",
+    "token_embd.weight": "q4_k"
+  },
+  "bytes": {
+    "heuristic_body+embd": 4880019456,
+    "map_body+embd": 4920848384,
+    "ratio": 1.0084,
+    "note": "quantizable, unpinned tensors only; pinned/unquantized tensors are identical in both"
+  },
+  "body_base_share": 0.9,
+  "body_bpw": 4.677,
+  "body_bpw_heuristic": 4.63,
+  "total_params": 8190735360,
+  "file_bytes_pred": 4922081280,
+  "file_bpw_pred": 4.807,
+  "histogram": {
+    "attn_k": {
+      "q6_k": 26,
+      "q4_k": 9,
+      "q8_0": 1
+    },
+    "attn_output": {
+      "q4_k": 28,
+      "q6_k": 8
+    },
+    "attn_q": {
+      "q4_k": 36
+    },
+    "attn_v": {
+      "q8_0": 36
+    },
+    "ffn_down": {
+      "q4_k": 36
+    },
+    "ffn_gate": {
+      "q4_k": 36
+    },
+    "ffn_up": {
+      "q4_k": 35,
+      "q6_k": 1
+    }
+  },
+  "prior": "layout/prior.json",
+  "pin_rule": "floor",
+  "embd_cap": "heur",
+  "generator": "auto_quant_v2.layout",
+  "generator_key": "3bf8b43e20a68f0b",
+  "quant_type": "Q4_K_S",
+  "quantize_ftype": "Q4_K_S",
+  "llama_cpp_version": "b11490",
+  "unknown_kinds": {},
+  "commit": "5d550ef5561a30307f1d436db2c6835b75395e77"
+}

layouts/AesCode-8B-Q4_K_S.tensor-types.txt

@@ -0,0 +1,254 @@
+^output\.weight$=q6_k
+^token_embd\.weight$=q4_k
+^blk\.0\.attn_k\.weight$=q8_0
+^blk\.0\.attn_output\.weight$=q4_k
+^blk\.0\.attn_q\.weight$=q4_k
+^blk\.0\.attn_v\.weight$=q8_0
+^blk\.0\.ffn_down\.weight$=q4_k
+^blk\.0\.ffn_gate\.weight$=q4_k
+^blk\.0\.ffn_up\.weight$=q4_k
+^blk\.1\.attn_k\.weight$=q6_k
+^blk\.1\.attn_output\.weight$=q6_k
+^blk\.1\.attn_q\.weight$=q4_k
+^blk\.1\.attn_v\.weight$=q8_0
+^blk\.1\.ffn_down\.weight$=q4_k
+^blk\.1\.ffn_gate\.weight$=q4_k
+^blk\.1\.ffn_up\.weight$=q4_k
+^blk\.2\.attn_k\.weight$=q6_k
+^blk\.2\.attn_output\.weight$=q6_k
+^blk\.2\.attn_q\.weight$=q4_k
+^blk\.2\.attn_v\.weight$=q8_0
+^blk\.2\.ffn_down\.weight$=q4_k
+^blk\.2\.ffn_gate\.weight$=q4_k
+^blk\.2\.ffn_up\.weight$=q4_k
+^blk\.3\.attn_k\.weight$=q6_k
+^blk\.3\.attn_output\.weight$=q6_k
+^blk\.3\.attn_q\.weight$=q4_k
+^blk\.3\.attn_v\.weight$=q8_0
+^blk\.3\.ffn_down\.weight$=q4_k
+^blk\.3\.ffn_gate\.weight$=q4_k
+^blk\.3\.ffn_up\.weight$=q4_k
+^blk\.4\.attn_k\.weight$=q6_k
+^blk\.4\.attn_output\.weight$=q6_k
+^blk\.4\.attn_q\.weight$=q4_k
+^blk\.4\.attn_v\.weight$=q8_0
+^blk\.4\.ffn_down\.weight$=q4_k
+^blk\.4\.ffn_gate\.weight$=q4_k
+^blk\.4\.ffn_up\.weight$=q4_k
+^blk\.5\.attn_k\.weight$=q6_k
+^blk\.5\.attn_output\.weight$=q6_k
+^blk\.5\.attn_q\.weight$=q4_k
+^blk\.5\.attn_v\.weight$=q8_0
+^blk\.5\.ffn_down\.weight$=q4_k
+^blk\.5\.ffn_gate\.weight$=q4_k
+^blk\.5\.ffn_up\.weight$=q4_k
+^blk\.6\.attn_k\.weight$=q6_k
+^blk\.6\.attn_output\.weight$=q6_k
+^blk\.6\.attn_q\.weight$=q4_k
+^blk\.6\.attn_v\.weight$=q8_0
+^blk\.6\.ffn_down\.weight$=q4_k
+^blk\.6\.ffn_gate\.weight$=q4_k
+^blk\.6\.ffn_up\.weight$=q4_k
+^blk\.7\.attn_k\.weight$=q6_k
+^blk\.7\.attn_output\.weight$=q6_k
+^blk\.7\.attn_q\.weight$=q4_k
+^blk\.7\.attn_v\.weight$=q8_0
+^blk\.7\.ffn_down\.weight$=q4_k
+^blk\.7\.ffn_gate\.weight$=q4_k
+^blk\.7\.ffn_up\.weight$=q4_k
+^blk\.8\.attn_k\.weight$=q6_k
+^blk\.8\.attn_output\.weight$=q6_k
+^blk\.8\.attn_q\.weight$=q4_k
+^blk\.8\.attn_v\.weight$=q8_0
+^blk\.8\.ffn_down\.weight$=q4_k
+^blk\.8\.ffn_gate\.weight$=q4_k
+^blk\.8\.ffn_up\.weight$=q4_k
+^blk\.9\.attn_k\.weight$=q6_k
+^blk\.9\.attn_output\.weight$=q4_k
+^blk\.9\.attn_q\.weight$=q4_k
+^blk\.9\.attn_v\.weight$=q8_0
+^blk\.9\.ffn_down\.weight$=q4_k
+^blk\.9\.ffn_gate\.weight$=q4_k
+^blk\.9\.ffn_up\.weight$=q4_k
+^blk\.10\.attn_k\.weight$=q6_k
+^blk\.10\.attn_output\.weight$=q4_k
+^blk\.10\.attn_q\.weight$=q4_k
+^blk\.10\.attn_v\.weight$=q8_0
+^blk\.10\.ffn_down\.weight$=q4_k
+^blk\.10\.ffn_gate\.weight$=q4_k
+^blk\.10\.ffn_up\.weight$=q4_k
+^blk\.11\.attn_k\.weight$=q6_k
+^blk\.11\.attn_output\.weight$=q4_k
+^blk\.11\.attn_q\.weight$=q4_k
+^blk\.11\.attn_v\.weight$=q8_0
+^blk\.11\.ffn_down\.weight$=q4_k
+^blk\.11\.ffn_gate\.weight$=q4_k
+^blk\.11\.ffn_up\.weight$=q4_k
+^blk\.12\.attn_k\.weight$=q6_k
+^blk\.12\.attn_output\.weight$=q4_k
+^blk\.12\.attn_q\.weight$=q4_k
+^blk\.12\.attn_v\.weight$=q8_0
+^blk\.12\.ffn_down\.weight$=q4_k
+^blk\.12\.ffn_gate\.weight$=q4_k
+^blk\.12\.ffn_up\.weight$=q4_k
+^blk\.13\.attn_k\.weight$=q6_k
+^blk\.13\.attn_output\.weight$=q4_k
+^blk\.13\.attn_q\.weight$=q4_k
+^blk\.13\.attn_v\.weight$=q8_0
+^blk\.13\.ffn_down\.weight$=q4_k
+^blk\.13\.ffn_gate\.weight$=q4_k
+^blk\.13\.ffn_up\.weight$=q4_k
+^blk\.14\.attn_k\.weight$=q6_k
+^blk\.14\.attn_output\.weight$=q4_k
+^blk\.14\.attn_q\.weight$=q4_k
+^blk\.14\.attn_v\.weight$=q8_0
+^blk\.14\.ffn_down\.weight$=q4_k
+^blk\.14\.ffn_gate\.weight$=q4_k
+^blk\.14\.ffn_up\.weight$=q4_k
+^blk\.15\.attn_k\.weight$=q6_k
+^blk\.15\.attn_output\.weight$=q4_k
+^blk\.15\.attn_q\.weight$=q4_k
+^blk\.15\.attn_v\.weight$=q8_0
+^blk\.15\.ffn_down\.weight$=q4_k
+^blk\.15\.ffn_gate\.weight$=q4_k
+^blk\.15\.ffn_up\.weight$=q4_k
+^blk\.16\.attn_k\.weight$=q6_k
+^blk\.16\.attn_output\.weight$=q4_k
+^blk\.16\.attn_q\.weight$=q4_k
+^blk\.16\.attn_v\.weight$=q8_0
+^blk\.16\.ffn_down\.weight$=q4_k
+^blk\.16\.ffn_gate\.weight$=q4_k
+^blk\.16\.ffn_up\.weight$=q4_k
+^blk\.17\.attn_k\.weight$=q6_k
+^blk\.17\.attn_output\.weight$=q4_k
+^blk\.17\.attn_q\.weight$=q4_k
+^blk\.17\.attn_v\.weight$=q8_0
+^blk\.17\.ffn_down\.weight$=q4_k
+^blk\.17\.ffn_gate\.weight$=q4_k
+^blk\.17\.ffn_up\.weight$=q4_k
+^blk\.18\.attn_k\.weight$=q6_k
+^blk\.18\.attn_output\.weight$=q4_k
+^blk\.18\.attn_q\.weight$=q4_k
+^blk\.18\.attn_v\.weight$=q8_0
+^blk\.18\.ffn_down\.weight$=q4_k
+^blk\.18\.ffn_gate\.weight$=q4_k
+^blk\.18\.ffn_up\.weight$=q4_k
+^blk\.19\.attn_k\.weight$=q6_k
+^blk\.19\.attn_output\.weight$=q4_k
+^blk\.19\.attn_q\.weight$=q4_k
+^blk\.19\.attn_v\.weight$=q8_0
+^blk\.19\.ffn_down\.weight$=q4_k
+^blk\.19\.ffn_gate\.weight$=q4_k
+^blk\.19\.ffn_up\.weight$=q4_k
+^blk\.20\.attn_k\.weight$=q6_k
+^blk\.20\.attn_output\.weight$=q4_k
+^blk\.20\.attn_q\.weight$=q4_k
+^blk\.20\.attn_v\.weight$=q8_0
+^blk\.20\.ffn_down\.weight$=q4_k
+^blk\.20\.ffn_gate\.weight$=q4_k
+^blk\.20\.ffn_up\.weight$=q4_k
+^blk\.21\.attn_k\.weight$=q6_k
+^blk\.21\.attn_output\.weight$=q4_k
+^blk\.21\.attn_q\.weight$=q4_k
+^blk\.21\.attn_v\.weight$=q8_0
+^blk\.21\.ffn_down\.weight$=q4_k
+^blk\.21\.ffn_gate\.weight$=q4_k
+^blk\.21\.ffn_up\.weight$=q4_k
+^blk\.22\.attn_k\.weight$=q6_k
+^blk\.22\.attn_output\.weight$=q4_k
+^blk\.22\.attn_q\.weight$=q4_k
+^blk\.22\.attn_v\.weight$=q8_0
+^blk\.22\.ffn_down\.weight$=q4_k
+^blk\.22\.ffn_gate\.weight$=q4_k
+^blk\.22\.ffn_up\.weight$=q4_k
+^blk\.23\.attn_k\.weight$=q6_k
+^blk\.23\.attn_output\.weight$=q4_k
+^blk\.23\.attn_q\.weight$=q4_k
+^blk\.23\.attn_v\.weight$=q8_0
+^blk\.23\.ffn_down\.weight$=q4_k
+^blk\.23\.ffn_gate\.weight$=q4_k
+^blk\.23\.ffn_up\.weight$=q4_k
+^blk\.24\.attn_k\.weight$=q6_k
+^blk\.24\.attn_output\.weight$=q4_k
+^blk\.24\.attn_q\.weight$=q4_k
+^blk\.24\.attn_v\.weight$=q8_0
+^blk\.24\.ffn_down\.weight$=q4_k
+^blk\.24\.ffn_gate\.weight$=q4_k
+^blk\.24\.ffn_up\.weight$=q4_k
+^blk\.25\.attn_k\.weight$=q6_k
+^blk\.25\.attn_output\.weight$=q4_k
+^blk\.25\.attn_q\.weight$=q4_k
+^blk\.25\.attn_v\.weight$=q8_0
+^blk\.25\.ffn_down\.weight$=q4_k
+^blk\.25\.ffn_gate\.weight$=q4_k
+^blk\.25\.ffn_up\.weight$=q4_k
+^blk\.26\.attn_k\.weight$=q6_k
+^blk\.26\.attn_output\.weight$=q4_k
+^blk\.26\.attn_q\.weight$=q4_k
+^blk\.26\.attn_v\.weight$=q8_0
+^blk\.26\.ffn_down\.weight$=q4_k
+^blk\.26\.ffn_gate\.weight$=q4_k
+^blk\.26\.ffn_up\.weight$=q4_k
+^blk\.27\.attn_k\.weight$=q4_k
+^blk\.27\.attn_output\.weight$=q4_k
+^blk\.27\.attn_q\.weight$=q4_k
+^blk\.27\.attn_v\.weight$=q8_0
+^blk\.27\.ffn_down\.weight$=q4_k
+^blk\.27\.ffn_gate\.weight$=q4_k
+^blk\.27\.ffn_up\.weight$=q4_k
+^blk\.28\.attn_k\.weight$=q4_k
+^blk\.28\.attn_output\.weight$=q4_k
+^blk\.28\.attn_q\.weight$=q4_k
+^blk\.28\.attn_v\.weight$=q8_0
+^blk\.28\.ffn_down\.weight$=q4_k
+^blk\.28\.ffn_gate\.weight$=q4_k
+^blk\.28\.ffn_up\.weight$=q4_k
+^blk\.29\.attn_k\.weight$=q4_k
+^blk\.29\.attn_output\.weight$=q4_k
+^blk\.29\.attn_q\.weight$=q4_k
+^blk\.29\.attn_v\.weight$=q8_0
+^blk\.29\.ffn_down\.weight$=q4_k
+^blk\.29\.ffn_gate\.weight$=q4_k
+^blk\.29\.ffn_up\.weight$=q4_k
+^blk\.30\.attn_k\.weight$=q4_k
+^blk\.30\.attn_output\.weight$=q4_k
+^blk\.30\.attn_q\.weight$=q4_k
+^blk\.30\.attn_v\.weight$=q8_0
+^blk\.30\.ffn_down\.weight$=q4_k
+^blk\.30\.ffn_gate\.weight$=q4_k
+^blk\.30\.ffn_up\.weight$=q4_k
+^blk\.31\.attn_k\.weight$=q4_k
+^blk\.31\.attn_output\.weight$=q4_k
+^blk\.31\.attn_q\.weight$=q4_k
+^blk\.31\.attn_v\.weight$=q8_0
+^blk\.31\.ffn_down\.weight$=q4_k
+^blk\.31\.ffn_gate\.weight$=q4_k
+^blk\.31\.ffn_up\.weight$=q4_k
+^blk\.32\.attn_k\.weight$=q4_k
+^blk\.32\.attn_output\.weight$=q4_k
+^blk\.32\.attn_q\.weight$=q4_k
+^blk\.32\.attn_v\.weight$=q8_0
+^blk\.32\.ffn_down\.weight$=q4_k
+^blk\.32\.ffn_gate\.weight$=q4_k
+^blk\.32\.ffn_up\.weight$=q4_k
+^blk\.33\.attn_k\.weight$=q4_k
+^blk\.33\.attn_output\.weight$=q4_k
+^blk\.33\.attn_q\.weight$=q4_k
+^blk\.33\.attn_v\.weight$=q8_0
+^blk\.33\.ffn_down\.weight$=q4_k
+^blk\.33\.ffn_gate\.weight$=q4_k
+^blk\.33\.ffn_up\.weight$=q4_k
+^blk\.34\.attn_k\.weight$=q4_k
+^blk\.34\.attn_output\.weight$=q4_k
+^blk\.34\.attn_q\.weight$=q4_k
+^blk\.34\.attn_v\.weight$=q8_0
+^blk\.34\.ffn_down\.weight$=q4_k
+^blk\.34\.ffn_gate\.weight$=q4_k
+^blk\.34\.ffn_up\.weight$=q4_k
+^blk\.35\.attn_k\.weight$=q4_k
+^blk\.35\.attn_output\.weight$=q4_k
+^blk\.35\.attn_q\.weight$=q4_k
+^blk\.35\.attn_v\.weight$=q8_0
+^blk\.35\.ffn_down\.weight$=q4_k
+^blk\.35\.ffn_gate\.weight$=q4_k
+^blk\.35\.ffn_up\.weight$=q6_k

layouts/AesCode-8B-Q5_K_M.layout.json

@@ -0,0 +1,86 @@
+{
+  "model": "/media/m2teamgroup/models/AesCode-8B-GGUF/AesCode-8B-bf16.gguf",
+  "imatrix": "/media/m2teamgroup/models/AesCode-8B-GGUF/AesCode-8B-imatrix.gguf",
+  "ftype": "Q5_K_M",
+  "base_type": "q5_k",
+  "family": "k",
+  "recipe": "ident",
+  "variant": "shr0.70",
+  "auto_crush": false,
+  "rung_mode": "share",
+  "tier": "M",
+  "rung_share": 0.7,
+  "crush_pass_bytes": null,
+  "embd_shares": {
+    "output.weight": {
+      "share": 0.0853,
+      "type": "q6_k"
+    },
+    "token_embd.weight": {
+      "share": 0.0715,
+      "type": "q5_k"
+    }
+  },
+  "arch": "dense",
+  "tied": false,
+  "n_layer": 36,
+  "body_rule": true,
+  "pins": {},
+  "shape_pins": {},
+  "fixed_pins": {},
+  "embeddings": {
+    "output.weight": "q6_k",
+    "token_embd.weight": "q5_k"
+  },
+  "bytes": {
+    "heuristic_body+embd": 5984301056,
+    "map_body+embd": 6235598848,
+    "ratio": 1.042,
+    "note": "quantizable, unpinned tensors only; pinned/unquantized tensors are identical in both"
+  },
+  "body_base_share": 0.7,
+  "body_bpw": 6.101,
+  "body_bpw_heuristic": 5.812,
+  "total_params": 8190735360,
+  "file_bytes_pred": 6236831744,
+  "file_bpw_pred": 6.092,
+  "histogram": {
+    "attn_k": {
+      "q8_0": 27,
+      "q5_k": 6,
+      "q6_k": 3
+    },
+    "attn_output": {
+      "q5_k": 14,
+      "q6_k": 14,
+      "q8_0": 8
+    },
+    "attn_q": {
+      "q5_k": 36
+    },
+    "attn_v": {
+      "q8_0": 36
+    },
+    "ffn_down": {
+      "q5_k": 27,
+      "q8_0": 9
+    },
+    "ffn_gate": {
+      "q5_k": 36
+    },
+    "ffn_up": {
+      "q5_k": 27,
+      "q8_0": 9
+    }
+  },
+  "prior": "layout/prior.json",
+  "pin_rule": "floor",
+  "embd_cap": "heur",
+  "generator": "auto_quant_v2.layout",
+  "generator_key": "3bf8b43e20a68f0b",
+  "quant_type": "Q5_K_M",
+  "quantize_ftype": "Q5_K_M",
+  "llama_cpp_version": "b11490",
+  "unknown_kinds": {},
+  "commit": "5d550ef5561a30307f1d436db2c6835b75395e77"
+}

layouts/AesCode-8B-Q5_K_M.tensor-types.txt

@@ -0,0 +1,254 @@
+^output\.weight$=q6_k
+^token_embd\.weight$=q5_k
+^blk\.0\.attn_k\.weight$=q8_0
+^blk\.0\.attn_output\.weight$=q5_k
+^blk\.0\.attn_q\.weight$=q5_k
+^blk\.0\.attn_v\.weight$=q8_0
+^blk\.0\.ffn_down\.weight$=q5_k
+^blk\.0\.ffn_gate\.weight$=q5_k
+^blk\.0\.ffn_up\.weight$=q5_k
+^blk\.1\.attn_k\.weight$=q8_0
+^blk\.1\.attn_output\.weight$=q8_0
+^blk\.1\.attn_q\.weight$=q5_k
+^blk\.1\.attn_v\.weight$=q8_0
+^blk\.1\.ffn_down\.weight$=q8_0
+^blk\.1\.ffn_gate\.weight$=q5_k
+^blk\.1\.ffn_up\.weight$=q8_0
+^blk\.2\.attn_k\.weight$=q8_0
+^blk\.2\.attn_output\.weight$=q8_0
+^blk\.2\.attn_q\.weight$=q5_k
+^blk\.2\.attn_v\.weight$=q8_0
+^blk\.2\.ffn_down\.weight$=q8_0
+^blk\.2\.ffn_gate\.weight$=q5_k
+^blk\.2\.ffn_up\.weight$=q8_0
+^blk\.3\.attn_k\.weight$=q8_0
+^blk\.3\.attn_output\.weight$=q8_0
+^blk\.3\.attn_q\.weight$=q5_k
+^blk\.3\.attn_v\.weight$=q8_0
+^blk\.3\.ffn_down\.weight$=q8_0
+^blk\.3\.ffn_gate\.weight$=q5_k
+^blk\.3\.ffn_up\.weight$=q8_0
+^blk\.4\.attn_k\.weight$=q8_0
+^blk\.4\.attn_output\.weight$=q8_0
+^blk\.4\.attn_q\.weight$=q5_k
+^blk\.4\.attn_v\.weight$=q8_0
+^blk\.4\.ffn_down\.weight$=q8_0
+^blk\.4\.ffn_gate\.weight$=q5_k
+^blk\.4\.ffn_up\.weight$=q8_0
+^blk\.5\.attn_k\.weight$=q8_0
+^blk\.5\.attn_output\.weight$=q8_0
+^blk\.5\.attn_q\.weight$=q5_k
+^blk\.5\.attn_v\.weight$=q8_0
+^blk\.5\.ffn_down\.weight$=q8_0
+^blk\.5\.ffn_gate\.weight$=q5_k
+^blk\.5\.ffn_up\.weight$=q8_0
+^blk\.6\.attn_k\.weight$=q8_0
+^blk\.6\.attn_output\.weight$=q8_0
+^blk\.6\.attn_q\.weight$=q5_k
+^blk\.6\.attn_v\.weight$=q8_0
+^blk\.6\.ffn_down\.weight$=q8_0
+^blk\.6\.ffn_gate\.weight$=q5_k
+^blk\.6\.ffn_up\.weight$=q8_0
+^blk\.7\.attn_k\.weight$=q8_0
+^blk\.7\.attn_output\.weight$=q8_0
+^blk\.7\.attn_q\.weight$=q5_k
+^blk\.7\.attn_v\.weight$=q8_0
+^blk\.7\.ffn_down\.weight$=q8_0
+^blk\.7\.ffn_gate\.weight$=q5_k
+^blk\.7\.ffn_up\.weight$=q8_0
+^blk\.8\.attn_k\.weight$=q8_0
+^blk\.8\.attn_output\.weight$=q8_0
+^blk\.8\.attn_q\.weight$=q5_k
+^blk\.8\.attn_v\.weight$=q8_0
+^blk\.8\.ffn_down\.weight$=q8_0
+^blk\.8\.ffn_gate\.weight$=q5_k
+^blk\.8\.ffn_up\.weight$=q8_0
+^blk\.9\.attn_k\.weight$=q8_0
+^blk\.9\.attn_output\.weight$=q6_k
+^blk\.9\.attn_q\.weight$=q5_k
+^blk\.9\.attn_v\.weight$=q8_0
+^blk\.9\.ffn_down\.weight$=q5_k
+^blk\.9\.ffn_gate\.weight$=q5_k
+^blk\.9\.ffn_up\.weight$=q5_k
+^blk\.10\.attn_k\.weight$=q8_0
+^blk\.10\.attn_output\.weight$=q6_k
+^blk\.10\.attn_q\.weight$=q5_k
+^blk\.10\.attn_v\.weight$=q8_0
+^blk\.10\.ffn_down\.weight$=q5_k
+^blk\.10\.ffn_gate\.weight$=q5_k
+^blk\.10\.ffn_up\.weight$=q5_k
+^blk\.11\.attn_k\.weight$=q8_0
+^blk\.11\.attn_output\.weight$=q6_k
+^blk\.11\.attn_q\.weight$=q5_k
+^blk\.11\.attn_v\.weight$=q8_0
+^blk\.11\.ffn_down\.weight$=q5_k
+^blk\.11\.ffn_gate\.weight$=q5_k
+^blk\.11\.ffn_up\.weight$=q5_k
+^blk\.12\.attn_k\.weight$=q8_0
+^blk\.12\.attn_output\.weight$=q6_k
+^blk\.12\.attn_q\.weight$=q5_k
+^blk\.12\.attn_v\.weight$=q8_0
+^blk\.12\.ffn_down\.weight$=q5_k
+^blk\.12\.ffn_gate\.weight$=q5_k
+^blk\.12\.ffn_up\.weight$=q5_k
+^blk\.13\.attn_k\.weight$=q8_0
+^blk\.13\.attn_output\.weight$=q6_k
+^blk\.13\.attn_q\.weight$=q5_k
+^blk\.13\.attn_v\.weight$=q8_0
+^blk\.13\.ffn_down\.weight$=q5_k
+^blk\.13\.ffn_gate\.weight$=q5_k
+^blk\.13\.ffn_up\.weight$=q5_k
+^blk\.14\.attn_k\.weight$=q8_0
+^blk\.14\.attn_output\.weight$=q6_k
+^blk\.14\.attn_q\.weight$=q5_k
+^blk\.14\.attn_v\.weight$=q8_0
+^blk\.14\.ffn_down\.weight$=q5_k
+^blk\.14\.ffn_gate\.weight$=q5_k
+^blk\.14\.ffn_up\.weight$=q5_k
+^blk\.15\.attn_k\.weight$=q8_0
+^blk\.15\.attn_output\.weight$=q6_k
+^blk\.15\.attn_q\.weight$=q5_k
+^blk\.15\.attn_v\.weight$=q8_0
+^blk\.15\.ffn_down\.weight$=q5_k
+^blk\.15\.ffn_gate\.weight$=q5_k
+^blk\.15\.ffn_up\.weight$=q5_k
+^blk\.16\.attn_k\.weight$=q8_0
+^blk\.16\.attn_output\.weight$=q6_k
+^blk\.16\.attn_q\.weight$=q5_k
+^blk\.16\.attn_v\.weight$=q8_0
+^blk\.16\.ffn_down\.weight$=q5_k
+^blk\.16\.ffn_gate\.weight$=q5_k
+^blk\.16\.ffn_up\.weight$=q5_k
+^blk\.17\.attn_k\.weight$=q8_0
+^blk\.17\.attn_output\.weight$=q6_k
+^blk\.17\.attn_q\.weight$=q5_k
+^blk\.17\.attn_v\.weight$=q8_0
+^blk\.17\.ffn_down\.weight$=q5_k
+^blk\.17\.ffn_gate\.weight$=q5_k
+^blk\.17\.ffn_up\.weight$=q5_k
+^blk\.18\.attn_k\.weight$=q8_0
+^blk\.18\.attn_output\.weight$=q6_k
+^blk\.18\.attn_q\.weight$=q5_k
+^blk\.18\.attn_v\.weight$=q8_0
+^blk\.18\.ffn_down\.weight$=q5_k
+^blk\.18\.ffn_gate\.weight$=q5_k
+^blk\.18\.ffn_up\.weight$=q5_k
+^blk\.19\.attn_k\.weight$=q8_0
+^blk\.19\.attn_output\.weight$=q6_k
+^blk\.19\.attn_q\.weight$=q5_k
+^blk\.19\.attn_v\.weight$=q8_0
+^blk\.19\.ffn_down\.weight$=q5_k
+^blk\.19\.ffn_gate\.weight$=q5_k
+^blk\.19\.ffn_up\.weight$=q5_k
+^blk\.20\.attn_k\.weight$=q8_0
+^blk\.20\.attn_output\.weight$=q6_k
+^blk\.20\.attn_q\.weight$=q5_k
+^blk\.20\.attn_v\.weight$=q8_0
+^blk\.20\.ffn_down\.weight$=q5_k
+^blk\.20\.ffn_gate\.weight$=q5_k
+^blk\.20\.ffn_up\.weight$=q5_k
+^blk\.21\.attn_k\.weight$=q8_0
+^blk\.21\.attn_output\.weight$=q6_k
+^blk\.21\.attn_q\.weight$=q5_k
+^blk\.21\.attn_v\.weight$=q8_0
+^blk\.21\.ffn_down\.weight$=q5_k
+^blk\.21\.ffn_gate\.weight$=q5_k
+^blk\.21\.ffn_up\.weight$=q5_k
+^blk\.22\.attn_k\.weight$=q8_0
+^blk\.22\.attn_output\.weight$=q6_k
+^blk\.22\.attn_q\.weight$=q5_k
+^blk\.22\.attn_v\.weight$=q8_0
+^blk\.22\.ffn_down\.weight$=q5_k
+^blk\.22\.ffn_gate\.weight$=q5_k
+^blk\.22\.ffn_up\.weight$=q5_k
+^blk\.23\.attn_k\.weight$=q8_0
+^blk\.23\.attn_output\.weight$=q5_k
+^blk\.23\.attn_q\.weight$=q5_k
+^blk\.23\.attn_v\.weight$=q8_0
+^blk\.23\.ffn_down\.weight$=q5_k
+^blk\.23\.ffn_gate\.weight$=q5_k
+^blk\.23\.ffn_up\.weight$=q5_k
+^blk\.24\.attn_k\.weight$=q8_0
+^blk\.24\.attn_output\.weight$=q5_k
+^blk\.24\.attn_q\.weight$=q5_k
+^blk\.24\.attn_v\.weight$=q8_0
+^blk\.24\.ffn_down\.weight$=q5_k
+^blk\.24\.ffn_gate\.weight$=q5_k
+^blk\.24\.ffn_up\.weight$=q5_k
+^blk\.25\.attn_k\.weight$=q8_0
+^blk\.25\.attn_output\.weight$=q5_k
+^blk\.25\.attn_q\.weight$=q5_k
+^blk\.25\.attn_v\.weight$=q8_0
+^blk\.25\.ffn_down\.weight$=q5_k
+^blk\.25\.ffn_gate\.weight$=q5_k
+^blk\.25\.ffn_up\.weight$=q5_k
+^blk\.26\.attn_k\.weight$=q8_0
+^blk\.26\.attn_output\.weight$=q5_k
+^blk\.26\.attn_q\.weight$=q5_k
+^blk\.26\.attn_v\.weight$=q8_0
+^blk\.26\.ffn_down\.weight$=q5_k
+^blk\.26\.ffn_gate\.weight$=q5_k
+^blk\.26\.ffn_up\.weight$=q5_k
+^blk\.27\.attn_k\.weight$=q6_k
+^blk\.27\.attn_output\.weight$=q5_k
+^blk\.27\.attn_q\.weight$=q5_k
+^blk\.27\.attn_v\.weight$=q8_0
+^blk\.27\.ffn_down\.weight$=q5_k
+^blk\.27\.ffn_gate\.weight$=q5_k
+^blk\.27\.ffn_up\.weight$=q5_k
+^blk\.28\.attn_k\.weight$=q6_k
+^blk\.28\.attn_output\.weight$=q5_k
+^blk\.28\.attn_q\.weight$=q5_k
+^blk\.28\.attn_v\.weight$=q8_0
+^blk\.28\.ffn_down\.weight$=q5_k
+^blk\.28\.ffn_gate\.weight$=q5_k
+^blk\.28\.ffn_up\.weight$=q5_k
+^blk\.29\.attn_k\.weight$=q6_k
+^blk\.29\.attn_output\.weight$=q5_k
+^blk\.29\.attn_q\.weight$=q5_k
+^blk\.29\.attn_v\.weight$=q8_0
+^blk\.29\.ffn_down\.weight$=q5_k
+^blk\.29\.ffn_gate\.weight$=q5_k
+^blk\.29\.ffn_up\.weight$=q5_k
+^blk\.30\.attn_k\.weight$=q5_k
+^blk\.30\.attn_output\.weight$=q5_k
+^blk\.30\.attn_q\.weight$=q5_k
+^blk\.30\.attn_v\.weight$=q8_0
+^blk\.30\.ffn_down\.weight$=q5_k
+^blk\.30\.ffn_gate\.weight$=q5_k
+^blk\.30\.ffn_up\.weight$=q5_k
+^blk\.31\.attn_k\.weight$=q5_k
+^blk\.31\.attn_output\.weight$=q5_k
+^blk\.31\.attn_q\.weight$=q5_k
+^blk\.31\.attn_v\.weight$=q8_0
+^blk\.31\.ffn_down\.weight$=q5_k
+^blk\.31\.ffn_gate\.weight$=q5_k
+^blk\.31\.ffn_up\.weight$=q5_k
+^blk\.32\.attn_k\.weight$=q5_k
+^blk\.32\.attn_output\.weight$=q5_k
+^blk\.32\.attn_q\.weight$=q5_k
+^blk\.32\.attn_v\.weight$=q8_0
+^blk\.32\.ffn_down\.weight$=q5_k
+^blk\.32\.ffn_gate\.weight$=q5_k
+^blk\.32\.ffn_up\.weight$=q5_k
+^blk\.33\.attn_k\.weight$=q5_k
+^blk\.33\.attn_output\.weight$=q5_k
+^blk\.33\.attn_q\.weight$=q5_k
+^blk\.33\.attn_v\.weight$=q8_0
+^blk\.33\.ffn_down\.weight$=q5_k
+^blk\.33\.ffn_gate\.weight$=q5_k
+^blk\.33\.ffn_up\.weight$=q5_k
+^blk\.34\.attn_k\.weight$=q5_k
+^blk\.34\.attn_output\.weight$=q5_k
+^blk\.34\.attn_q\.weight$=q5_k
+^blk\.34\.attn_v\.weight$=q8_0
+^blk\.34\.ffn_down\.weight$=q5_k
+^blk\.34\.ffn_gate\.weight$=q5_k
+^blk\.34\.ffn_up\.weight$=q5_k
+^blk\.35\.attn_k\.weight$=q5_k
+^blk\.35\.attn_output\.weight$=q5_k
+^blk\.35\.attn_q\.weight$=q5_k
+^blk\.35\.attn_v\.weight$=q8_0
+^blk\.35\.ffn_down\.weight$=q8_0
+^blk\.35\.ffn_gate\.weight$=q5_k
+^blk\.35\.ffn_up\.weight$=q8_0

layouts/AesCode-8B-Q5_K_S.layout.json

@@ -0,0 +1,85 @@
+{
+  "model": "/media/m2teamgroup/models/AesCode-8B-GGUF/AesCode-8B-bf16.gguf",
+  "imatrix": "/media/m2teamgroup/models/AesCode-8B-GGUF/AesCode-8B-imatrix.gguf",
+  "ftype": "Q5_K_S",
+  "base_type": "q5_k",
+  "family": "k",
+  "recipe": "ident",
+  "variant": "shr0.90",
+  "auto_crush": false,
+  "rung_mode": "share",
+  "tier": "S",
+  "rung_share": 0.9,
+  "crush_pass_bytes": null,
+  "embd_shares": {
+    "output.weight": {
+      "share": 0.0887,
+      "type": "q6_k"
+    },
+    "token_embd.weight": {
+      "share": 0.0743,
+      "type": "q5_k"
+    }
+  },
+  "arch": "dense",
+  "tied": false,
+  "n_layer": 36,
+  "body_rule": true,
+  "pins": {},
+  "shape_pins": {},
+  "fixed_pins": {},
+  "embeddings": {
+    "output.weight": "q6_k",
+    "token_embd.weight": "q5_k"
+  },
+  "bytes": {
+    "heuristic_body+embd": 5753679872,
+    "map_body+embd": 5884653568,
+    "ratio": 1.0228,
+    "note": "quantizable, unpinned tensors only; pinned/unquantized tensors are identical in both"
+  },
+  "body_base_share": 0.9,
+  "body_bpw": 5.697,
+  "body_bpw_heuristic": 5.546,
+  "total_params": 8190735360,
+  "file_bytes_pred": 5885886464,
+  "file_bpw_pred": 5.749,
+  "histogram": {
+    "attn_k": {
+      "q8_0": 27,
+      "q5_k": 8,
+      "q6_k": 1
+    },
+    "attn_output": {
+      "q5_k": 27,
+      "q8_0": 8,
+      "q6_k": 1
+    },
+    "attn_q": {
+      "q5_k": 36
+    },
+    "attn_v": {
+      "q8_0": 36
+    },
+    "ffn_down": {
+      "q5_k": 36
+    },
+    "ffn_gate": {
+      "q5_k": 36
+    },
+    "ffn_up": {
+      "q5_k": 35,
+      "q8_0": 1
+    }
+  },
+  "prior": "layout/prior.json",
+  "pin_rule": "floor",
+  "embd_cap": "heur",
+  "generator": "auto_quant_v2.layout",
+  "generator_key": "3bf8b43e20a68f0b",
+  "quant_type": "Q5_K_S",
+  "quantize_ftype": "Q5_K_S",
+  "llama_cpp_version": "b11490",
+  "unknown_kinds": {},
+  "commit": "5d550ef5561a30307f1d436db2c6835b75395e77"
+}

layouts/AesCode-8B-Q5_K_S.tensor-types.txt

@@ -0,0 +1,254 @@
+^output\.weight$=q6_k
+^token_embd\.weight$=q5_k
+^blk\.0\.attn_k\.weight$=q8_0
+^blk\.0\.attn_output\.weight$=q5_k
+^blk\.0\.attn_q\.weight$=q5_k
+^blk\.0\.attn_v\.weight$=q8_0
+^blk\.0\.ffn_down\.weight$=q5_k
+^blk\.0\.ffn_gate\.weight$=q5_k
+^blk\.0\.ffn_up\.weight$=q5_k
+^blk\.1\.attn_k\.weight$=q8_0
+^blk\.1\.attn_output\.weight$=q8_0
+^blk\.1\.attn_q\.weight$=q5_k
+^blk\.1\.attn_v\.weight$=q8_0
+^blk\.1\.ffn_down\.weight$=q5_k
+^blk\.1\.ffn_gate\.weight$=q5_k
+^blk\.1\.ffn_up\.weight$=q5_k
+^blk\.2\.attn_k\.weight$=q8_0
+^blk\.2\.attn_output\.weight$=q8_0
+^blk\.2\.attn_q\.weight$=q5_k
+^blk\.2\.attn_v\.weight$=q8_0
+^blk\.2\.ffn_down\.weight$=q5_k
+^blk\.2\.ffn_gate\.weight$=q5_k
+^blk\.2\.ffn_up\.weight$=q5_k
+^blk\.3\.attn_k\.weight$=q8_0
+^blk\.3\.attn_output\.weight$=q8_0
+^blk\.3\.attn_q\.weight$=q5_k
+^blk\.3\.attn_v\.weight$=q8_0
+^blk\.3\.ffn_down\.weight$=q5_k
+^blk\.3\.ffn_gate\.weight$=q5_k
+^blk\.3\.ffn_up\.weight$=q5_k
+^blk\.4\.attn_k\.weight$=q8_0
+^blk\.4\.attn_output\.weight$=q8_0
+^blk\.4\.attn_q\.weight$=q5_k
+^blk\.4\.attn_v\.weight$=q8_0
+^blk\.4\.ffn_down\.weight$=q5_k
+^blk\.4\.ffn_gate\.weight$=q5_k
+^blk\.4\.ffn_up\.weight$=q5_k
+^blk\.5\.attn_k\.weight$=q8_0
+^blk\.5\.attn_output\.weight$=q8_0
+^blk\.5\.attn_q\.weight$=q5_k
+^blk\.5\.attn_v\.weight$=q8_0
+^blk\.5\.ffn_down\.weight$=q5_k
+^blk\.5\.ffn_gate\.weight$=q5_k
+^blk\.5\.ffn_up\.weight$=q5_k
+^blk\.6\.attn_k\.weight$=q8_0
+^blk\.6\.attn_output\.weight$=q8_0
+^blk\.6\.attn_q\.weight$=q5_k
+^blk\.6\.attn_v\.weight$=q8_0
+^blk\.6\.ffn_down\.weight$=q5_k
+^blk\.6\.ffn_gate\.weight$=q5_k
+^blk\.6\.ffn_up\.weight$=q5_k
+^blk\.7\.attn_k\.weight$=q8_0
+^blk\.7\.attn_output\.weight$=q8_0
+^blk\.7\.attn_q\.weight$=q5_k
+^blk\.7\.attn_v\.weight$=q8_0
+^blk\.7\.ffn_down\.weight$=q5_k
+^blk\.7\.ffn_gate\.weight$=q5_k
+^blk\.7\.ffn_up\.weight$=q5_k
+^blk\.8\.attn_k\.weight$=q8_0
+^blk\.8\.attn_output\.weight$=q8_0
+^blk\.8\.attn_q\.weight$=q5_k
+^blk\.8\.attn_v\.weight$=q8_0
+^blk\.8\.ffn_down\.weight$=q5_k
+^blk\.8\.ffn_gate\.weight$=q5_k
+^blk\.8\.ffn_up\.weight$=q5_k
+^blk\.9\.attn_k\.weight$=q8_0
+^blk\.9\.attn_output\.weight$=q6_k
+^blk\.9\.attn_q\.weight$=q5_k
+^blk\.9\.attn_v\.weight$=q8_0
+^blk\.9\.ffn_down\.weight$=q5_k
+^blk\.9\.ffn_gate\.weight$=q5_k
+^blk\.9\.ffn_up\.weight$=q5_k
+^blk\.10\.attn_k\.weight$=q8_0
+^blk\.10\.attn_output\.weight$=q5_k
+^blk\.10\.attn_q\.weight$=q5_k
+^blk\.10\.attn_v\.weight$=q8_0
+^blk\.10\.ffn_down\.weight$=q5_k
+^blk\.10\.ffn_gate\.weight$=q5_k
+^blk\.10\.ffn_up\.weight$=q5_k
+^blk\.11\.attn_k\.weight$=q8_0
+^blk\.11\.attn_output\.weight$=q5_k
+^blk\.11\.attn_q\.weight$=q5_k
+^blk\.11\.attn_v\.weight$=q8_0
+^blk\.11\.ffn_down\.weight$=q5_k
+^blk\.11\.ffn_gate\.weight$=q5_k
+^blk\.11\.ffn_up\.weight$=q5_k
+^blk\.12\.attn_k\.weight$=q8_0
+^blk\.12\.attn_output\.weight$=q5_k
+^blk\.12\.attn_q\.weight$=q5_k
+^blk\.12\.attn_v\.weight$=q8_0
+^blk\.12\.ffn_down\.weight$=q5_k
+^blk\.12\.ffn_gate\.weight$=q5_k
+^blk\.12\.ffn_up\.weight$=q5_k
+^blk\.13\.attn_k\.weight$=q8_0
+^blk\.13\.attn_output\.weight$=q5_k
+^blk\.13\.attn_q\.weight$=q5_k
+^blk\.13\.attn_v\.weight$=q8_0
+^blk\.13\.ffn_down\.weight$=q5_k
+^blk\.13\.ffn_gate\.weight$=q5_k
+^blk\.13\.ffn_up\.weight$=q5_k
+^blk\.14\.attn_k\.weight$=q8_0
+^blk\.14\.attn_output\.weight$=q5_k
+^blk\.14\.attn_q\.weight$=q5_k
+^blk\.14\.attn_v\.weight$=q8_0
+^blk\.14\.ffn_down\.weight$=q5_k
+^blk\.14\.ffn_gate\.weight$=q5_k
+^blk\.14\.ffn_up\.weight$=q5_k
+^blk\.15\.attn_k\.weight$=q8_0
+^blk\.15\.attn_output\.weight$=q5_k
+^blk\.15\.attn_q\.weight$=q5_k
+^blk\.15\.attn_v\.weight$=q8_0
+^blk\.15\.ffn_down\.weight$=q5_k
+^blk\.15\.ffn_gate\.weight$=q5_k
+^blk\.15\.ffn_up\.weight$=q5_k
+^blk\.16\.attn_k\.weight$=q8_0
+^blk\.16\.attn_output\.weight$=q5_k
+^blk\.16\.attn_q\.weight$=q5_k
+^blk\.16\.attn_v\.weight$=q8_0
+^blk\.16\.ffn_down\.weight$=q5_k
+^blk\.16\.ffn_gate\.weight$=q5_k
+^blk\.16\.ffn_up\.weight$=q5_k
+^blk\.17\.attn_k\.weight$=q8_0
+^blk\.17\.attn_output\.weight$=q5_k
+^blk\.17\.attn_q\.weight$=q5_k
+^blk\.17\.attn_v\.weight$=q8_0
+^blk\.17\.ffn_down\.weight$=q5_k
+^blk\.17\.ffn_gate\.weight$=q5_k
+^blk\.17\.ffn_up\.weight$=q5_k
+^blk\.18\.attn_k\.weight$=q8_0
+^blk\.18\.attn_output\.weight$=q5_k
+^blk\.18\.attn_q\.weight$=q5_k
+^blk\.18\.attn_v\.weight$=q8_0
+^blk\.18\.ffn_down\.weight$=q5_k
+^blk\.18\.ffn_gate\.weight$=q5_k
+^blk\.18\.ffn_up\.weight$=q5_k
+^blk\.19\.attn_k\.weight$=q8_0
+^blk\.19\.attn_output\.weight$=q5_k
+^blk\.19\.attn_q\.weight$=q5_k
+^blk\.19\.attn_v\.weight$=q8_0
+^blk\.19\.ffn_down\.weight$=q5_k
+^blk\.19\.ffn_gate\.weight$=q5_k
+^blk\.19\.ffn_up\.weight$=q5_k
+^blk\.20\.attn_k\.weight$=q8_0
+^blk\.20\.attn_output\.weight$=q5_k
+^blk\.20\.attn_q\.weight$=q5_k
+^blk\.20\.attn_v\.weight$=q8_0
+^blk\.20\.ffn_down\.weight$=q5_k
+^blk\.20\.ffn_gate\.weight$=q5_k
+^blk\.20\.ffn_up\.weight$=q5_k
+^blk\.21\.attn_k\.weight$=q8_0
+^blk\.21\.attn_output\.weight$=q5_k
+^blk\.21\.attn_q\.weight$=q5_k
+^blk\.21\.attn_v\.weight$=q8_0
+^blk\.21\.ffn_down\.weight$=q5_k
+^blk\.21\.ffn_gate\.weight$=q5_k
+^blk\.21\.ffn_up\.weight$=q5_k
+^blk\.22\.attn_k\.weight$=q8_0
+^blk\.22\.attn_output\.weight$=q5_k
+^blk\.22\.attn_q\.weight$=q5_k
+^blk\.22\.attn_v\.weight$=q8_0
+^blk\.22\.ffn_down\.weight$=q5_k
+^blk\.22\.ffn_gate\.weight$=q5_k
+^blk\.22\.ffn_up\.weight$=q5_k
+^blk\.23\.attn_k\.weight$=q8_0
+^blk\.23\.attn_output\.weight$=q5_k
+^blk\.23\.attn_q\.weight$=q5_k
+^blk\.23\.attn_v\.weight$=q8_0
+^blk\.23\.ffn_down\.weight$=q5_k
+^blk\.23\.ffn_gate\.weight$=q5_k
+^blk\.23\.ffn_up\.weight$=q5_k
+^blk\.24\.attn_k\.weight$=q8_0
+^blk\.24\.attn_output\.weight$=q5_k
+^blk\.24\.attn_q\.weight$=q5_k
+^blk\.24\.attn_v\.weight$=q8_0
+^blk\.24\.ffn_down\.weight$=q5_k
+^blk\.24\.ffn_gate\.weight$=q5_k
+^blk\.24\.ffn_up\.weight$=q5_k
+^blk\.25\.attn_k\.weight$=q8_0
+^blk\.25\.attn_output\.weight$=q5_k
+^blk\.25\.attn_q\.weight$=q5_k
+^blk\.25\.attn_v\.weight$=q8_0
+^blk\.25\.ffn_down\.weight$=q5_k
+^blk\.25\.ffn_gate\.weight$=q5_k
+^blk\.25\.ffn_up\.weight$=q5_k
+^blk\.26\.attn_k\.weight$=q8_0
+^blk\.26\.attn_output\.weight$=q5_k
+^blk\.26\.attn_q\.weight$=q5_k
+^blk\.26\.attn_v\.weight$=q8_0
+^blk\.26\.ffn_down\.weight$=q5_k
+^blk\.26\.ffn_gate\.weight$=q5_k
+^blk\.26\.ffn_up\.weight$=q5_k
+^blk\.27\.attn_k\.weight$=q6_k
+^blk\.27\.attn_output\.weight$=q5_k
+^blk\.27\.attn_q\.weight$=q5_k
+^blk\.27\.attn_v\.weight$=q8_0
+^blk\.27\.ffn_down\.weight$=q5_k
+^blk\.27\.ffn_gate\.weight$=q5_k
+^blk\.27\.ffn_up\.weight$=q5_k
+^blk\.28\.attn_k\.weight$=q5_k
+^blk\.28\.attn_output\.weight$=q5_k
+^blk\.28\.attn_q\.weight$=q5_k
+^blk\.28\.attn_v\.weight$=q8_0
+^blk\.28\.ffn_down\.weight$=q5_k
+^blk\.28\.ffn_gate\.weight$=q5_k
+^blk\.28\.ffn_up\.weight$=q5_k
+^blk\.29\.attn_k\.weight$=q5_k
+^blk\.29\.attn_output\.weight$=q5_k
+^blk\.29\.attn_q\.weight$=q5_k
+^blk\.29\.attn_v\.weight$=q8_0
+^blk\.29\.ffn_down\.weight$=q5_k
+^blk\.29\.ffn_gate\.weight$=q5_k
+^blk\.29\.ffn_up\.weight$=q5_k
+^blk\.30\.attn_k\.weight$=q5_k
+^blk\.30\.attn_output\.weight$=q5_k
+^blk\.30\.attn_q\.weight$=q5_k
+^blk\.30\.attn_v\.weight$=q8_0
+^blk\.30\.ffn_down\.weight$=q5_k
+^blk\.30\.ffn_gate\.weight$=q5_k
+^blk\.30\.ffn_up\.weight$=q5_k
+^blk\.31\.attn_k\.weight$=q5_k
+^blk\.31\.attn_output\.weight$=q5_k
+^blk\.31\.attn_q\.weight$=q5_k
+^blk\.31\.attn_v\.weight$=q8_0
+^blk\.31\.ffn_down\.weight$=q5_k
+^blk\.31\.ffn_gate\.weight$=q5_k
+^blk\.31\.ffn_up\.weight$=q5_k
+^blk\.32\.attn_k\.weight$=q5_k
+^blk\.32\.attn_output\.weight$=q5_k
+^blk\.32\.attn_q\.weight$=q5_k
+^blk\.32\.attn_v\.weight$=q8_0
+^blk\.32\.ffn_down\.weight$=q5_k
+^blk\.32\.ffn_gate\.weight$=q5_k
+^blk\.32\.ffn_up\.weight$=q5_k
+^blk\.33\.attn_k\.weight$=q5_k
+^blk\.33\.attn_output\.weight$=q5_k
+^blk\.33\.attn_q\.weight$=q5_k
+^blk\.33\.attn_v\.weight$=q8_0
+^blk\.33\.ffn_down\.weight$=q5_k
+^blk\.33\.ffn_gate\.weight$=q5_k
+^blk\.33\.ffn_up\.weight$=q5_k
+^blk\.34\.attn_k\.weight$=q5_k
+^blk\.34\.attn_output\.weight$=q5_k
+^blk\.34\.attn_q\.weight$=q5_k
+^blk\.34\.attn_v\.weight$=q8_0
+^blk\.34\.ffn_down\.weight$=q5_k
+^blk\.34\.ffn_gate\.weight$=q5_k
+^blk\.34\.ffn_up\.weight$=q5_k
+^blk\.35\.attn_k\.weight$=q5_k
+^blk\.35\.attn_output\.weight$=q5_k
+^blk\.35\.attn_q\.weight$=q5_k
+^blk\.35\.attn_v\.weight$=q8_0
+^blk\.35\.ffn_down\.weight$=q5_k
+^blk\.35\.ffn_gate\.weight$=q5_k
+^blk\.35\.ffn_up\.weight$=q8_0

layouts/AesCode-8B-Q6_K.layout.json

@@ -0,0 +1,83 @@
+{
+  "model": "/media/m2teamgroup/models/AesCode-8B-GGUF/AesCode-8B-bf16.gguf",
+  "imatrix": "/media/m2teamgroup/models/AesCode-8B-GGUF/AesCode-8B-imatrix.gguf",
+  "ftype": "Q6_K",
+  "base_type": "q6_k",
+  "family": "k",
+  "recipe": "ident",
+  "variant": "shr0.70",
+  "auto_crush": false,
+  "rung_mode": "share",
+  "tier": "M",
+  "rung_share": 0.7,
+  "crush_pass_bytes": null,
+  "embd_shares": {
+    "output.weight": {
+      "share": 0.073,
+      "type": "q6_k"
+    },
+    "token_embd.weight": {
+      "share": 0.073,
+      "type": "q6_k"
+    }
+  },
+  "arch": "dense",
+  "tied": false,
+  "n_layer": 36,
+  "body_rule": true,
+  "pins": {},
+  "shape_pins": {},
+  "fixed_pins": {},
+  "embeddings": {
+    "output.weight": "q6_k",
+    "token_embd.weight": "q6_k"
+  },
+  "bytes": {
+    "heuristic_body+embd": 6992977920,
+    "map_body+embd": 7133159424,
+    "ratio": 1.02,
+    "note": "quantizable, unpinned tensors only; pinned/unquantized tensors are identical in both"
+  },
+  "body_base_share": 0.703,
+  "body_bpw": 7.04,
+  "body_bpw_heuristic": 6.878,
+  "total_params": 8190735360,
+  "file_bytes_pred": 7134392320,
+  "file_bpw_pred": 6.968,
+  "histogram": {
+    "attn_k": {
+      "q8_0": 36
+    },
+    "attn_output": {
+      "q8_0": 27,
+      "q6_k": 9
+    },
+    "attn_q": {
+      "q6_k": 36
+    },
+    "attn_v": {
+      "q8_0": 36
+    },
+    "ffn_down": {
+      "q6_k": 26,
+      "q8_0": 10
+    },
+    "ffn_gate": {
+      "q6_k": 36
+    },
+    "ffn_up": {
+      "q6_k": 27,
+      "q8_0": 9
+    }
+  },
+  "prior": "layout/prior.json",
+  "pin_rule": "floor",
+  "embd_cap": "heur",
+  "generator": "auto_quant_v2.layout",
+  "generator_key": "3bf8b43e20a68f0b",
+  "quant_type": "Q6_K",
+  "quantize_ftype": "Q6_K",
+  "llama_cpp_version": "b11490",
+  "unknown_kinds": {},
+  "commit": "5d550ef5561a30307f1d436db2c6835b75395e77"
+}

layouts/AesCode-8B-Q6_K.tensor-types.txt

@@ -0,0 +1,254 @@
+^output\.weight$=q6_k
+^token_embd\.weight$=q6_k
+^blk\.0\.attn_k\.weight$=q8_0
+^blk\.0\.attn_output\.weight$=q8_0
+^blk\.0\.attn_q\.weight$=q6_k
+^blk\.0\.attn_v\.weight$=q8_0
+^blk\.0\.ffn_down\.weight$=q8_0
+^blk\.0\.ffn_gate\.weight$=q6_k
+^blk\.0\.ffn_up\.weight$=q6_k
+^blk\.1\.attn_k\.weight$=q8_0
+^blk\.1\.attn_output\.weight$=q8_0
+^blk\.1\.attn_q\.weight$=q6_k
+^blk\.1\.attn_v\.weight$=q8_0
+^blk\.1\.ffn_down\.weight$=q8_0
+^blk\.1\.ffn_gate\.weight$=q6_k
+^blk\.1\.ffn_up\.weight$=q8_0
+^blk\.2\.attn_k\.weight$=q8_0
+^blk\.2\.attn_output\.weight$=q8_0
+^blk\.2\.attn_q\.weight$=q6_k
+^blk\.2\.attn_v\.weight$=q8_0
+^blk\.2\.ffn_down\.weight$=q8_0
+^blk\.2\.ffn_gate\.weight$=q6_k
+^blk\.2\.ffn_up\.weight$=q8_0
+^blk\.3\.attn_k\.weight$=q8_0
+^blk\.3\.attn_output\.weight$=q8_0
+^blk\.3\.attn_q\.weight$=q6_k
+^blk\.3\.attn_v\.weight$=q8_0
+^blk\.3\.ffn_down\.weight$=q8_0
+^blk\.3\.ffn_gate\.weight$=q6_k
+^blk\.3\.ffn_up\.weight$=q8_0
+^blk\.4\.attn_k\.weight$=q8_0
+^blk\.4\.attn_output\.weight$=q8_0
+^blk\.4\.attn_q\.weight$=q6_k
+^blk\.4\.attn_v\.weight$=q8_0
+^blk\.4\.ffn_down\.weight$=q8_0
+^blk\.4\.ffn_gate\.weight$=q6_k
+^blk\.4\.ffn_up\.weight$=q8_0
+^blk\.5\.attn_k\.weight$=q8_0
+^blk\.5\.attn_output\.weight$=q8_0
+^blk\.5\.attn_q\.weight$=q6_k
+^blk\.5\.attn_v\.weight$=q8_0
+^blk\.5\.ffn_down\.weight$=q8_0
+^blk\.5\.ffn_gate\.weight$=q6_k
+^blk\.5\.ffn_up\.weight$=q8_0
+^blk\.6\.attn_k\.weight$=q8_0
+^blk\.6\.attn_output\.weight$=q8_0
+^blk\.6\.attn_q\.weight$=q6_k
+^blk\.6\.attn_v\.weight$=q8_0
+^blk\.6\.ffn_down\.weight$=q8_0
+^blk\.6\.ffn_gate\.weight$=q6_k
+^blk\.6\.ffn_up\.weight$=q8_0
+^blk\.7\.attn_k\.weight$=q8_0
+^blk\.7\.attn_output\.weight$=q8_0
+^blk\.7\.attn_q\.weight$=q6_k
+^blk\.7\.attn_v\.weight$=q8_0
+^blk\.7\.ffn_down\.weight$=q8_0
+^blk\.7\.ffn_gate\.weight$=q6_k
+^blk\.7\.ffn_up\.weight$=q8_0
+^blk\.8\.attn_k\.weight$=q8_0
+^blk\.8\.attn_output\.weight$=q8_0
+^blk\.8\.attn_q\.weight$=q6_k
+^blk\.8\.attn_v\.weight$=q8_0
+^blk\.8\.ffn_down\.weight$=q8_0
+^blk\.8\.ffn_gate\.weight$=q6_k
+^blk\.8\.ffn_up\.weight$=q8_0
+^blk\.9\.attn_k\.weight$=q8_0
+^blk\.9\.attn_output\.weight$=q8_0
+^blk\.9\.attn_q\.weight$=q6_k
+^blk\.9\.attn_v\.weight$=q8_0
+^blk\.9\.ffn_down\.weight$=q6_k
+^blk\.9\.ffn_gate\.weight$=q6_k
+^blk\.9\.ffn_up\.weight$=q6_k
+^blk\.10\.attn_k\.weight$=q8_0
+^blk\.10\.attn_output\.weight$=q8_0
+^blk\.10\.attn_q\.weight$=q6_k
+^blk\.10\.attn_v\.weight$=q8_0
+^blk\.10\.ffn_down\.weight$=q6_k
+^blk\.10\.ffn_gate\.weight$=q6_k
+^blk\.10\.ffn_up\.weight$=q6_k
+^blk\.11\.attn_k\.weight$=q8_0
+^blk\.11\.attn_output\.weight$=q8_0
+^blk\.11\.attn_q\.weight$=q6_k
+^blk\.11\.attn_v\.weight$=q8_0
+^blk\.11\.ffn_down\.weight$=q6_k
+^blk\.11\.ffn_gate\.weight$=q6_k
+^blk\.11\.ffn_up\.weight$=q6_k
+^blk\.12\.attn_k\.weight$=q8_0
+^blk\.12\.attn_output\.weight$=q8_0
+^blk\.12\.attn_q\.weight$=q6_k
+^blk\.12\.attn_v\.weight$=q8_0
+^blk\.12\.ffn_down\.weight$=q6_k
+^blk\.12\.ffn_gate\.weight$=q6_k
+^blk\.12\.ffn_up\.weight$=q6_k
+^blk\.13\.attn_k\.weight$=q8_0
+^blk\.13\.attn_output\.weight$=q8_0
+^blk\.13\.attn_q\.weight$=q6_k
+^blk\.13\.attn_v\.weight$=q8_0
+^blk\.13\.ffn_down\.weight$=q6_k
+^blk\.13\.ffn_gate\.weight$=q6_k
+^blk\.13\.ffn_up\.weight$=q6_k
+^blk\.14\.attn_k\.weight$=q8_0
+^blk\.14\.attn_output\.weight$=q8_0
+^blk\.14\.attn_q\.weight$=q6_k
+^blk\.14\.attn_v\.weight$=q8_0
+^blk\.14\.ffn_down\.weight$=q6_k
+^blk\.14\.ffn_gate\.weight$=q6_k
+^blk\.14\.ffn_up\.weight$=q6_k
+^blk\.15\.attn_k\.weight$=q8_0
+^blk\.15\.attn_output\.weight$=q8_0
+^blk\.15\.attn_q\.weight$=q6_k
+^blk\.15\.attn_v\.weight$=q8_0
+^blk\.15\.ffn_down\.weight$=q6_k
+^blk\.15\.ffn_gate\.weight$=q6_k
+^blk\.15\.ffn_up\.weight$=q6_k
+^blk\.16\.attn_k\.weight$=q8_0
+^blk\.16\.attn_output\.weight$=q8_0
+^blk\.16\.attn_q\.weight$=q6_k
+^blk\.16\.attn_v\.weight$=q8_0
+^blk\.16\.ffn_down\.weight$=q6_k
+^blk\.16\.ffn_gate\.weight$=q6_k
+^blk\.16\.ffn_up\.weight$=q6_k
+^blk\.17\.attn_k\.weight$=q8_0
+^blk\.17\.attn_output\.weight$=q8_0
+^blk\.17\.attn_q\.weight$=q6_k
+^blk\.17\.attn_v\.weight$=q8_0
+^blk\.17\.ffn_down\.weight$=q6_k
+^blk\.17\.ffn_gate\.weight$=q6_k
+^blk\.17\.ffn_up\.weight$=q6_k
+^blk\.18\.attn_k\.weight$=q8_0
+^blk\.18\.attn_output\.weight$=q8_0
+^blk\.18\.attn_q\.weight$=q6_k
+^blk\.18\.attn_v\.weight$=q8_0
+^blk\.18\.ffn_down\.weight$=q6_k
+^blk\.18\.ffn_gate\.weight$=q6_k
+^blk\.18\.ffn_up\.weight$=q6_k
+^blk\.19\.attn_k\.weight$=q8_0
+^blk\.19\.attn_output\.weight$=q8_0
+^blk\.19\.attn_q\.weight$=q6_k
+^blk\.19\.attn_v\.weight$=q8_0
+^blk\.19\.ffn_down\.weight$=q6_k
+^blk\.19\.ffn_gate\.weight$=q6_k
+^blk\.19\.ffn_up\.weight$=q6_k
+^blk\.20\.attn_k\.weight$=q8_0
+^blk\.20\.attn_output\.weight$=q8_0
+^blk\.20\.attn_q\.weight$=q6_k
+^blk\.20\.attn_v\.weight$=q8_0
+^blk\.20\.ffn_down\.weight$=q6_k
+^blk\.20\.ffn_gate\.weight$=q6_k
+^blk\.20\.ffn_up\.weight$=q6_k
+^blk\.21\.attn_k\.weight$=q8_0
+^blk\.21\.attn_output\.weight$=q8_0
+^blk\.21\.attn_q\.weight$=q6_k
+^blk\.21\.attn_v\.weight$=q8_0
+^blk\.21\.ffn_down\.weight$=q6_k
+^blk\.21\.ffn_gate\.weight$=q6_k
+^blk\.21\.ffn_up\.weight$=q6_k
+^blk\.22\.attn_k\.weight$=q8_0
+^blk\.22\.attn_output\.weight$=q8_0
+^blk\.22\.attn_q\.weight$=q6_k
+^blk\.22\.attn_v\.weight$=q8_0
+^blk\.22\.ffn_down\.weight$=q6_k
+^blk\.22\.ffn_gate\.weight$=q6_k
+^blk\.22\.ffn_up\.weight$=q6_k
+^blk\.23\.attn_k\.weight$=q8_0
+^blk\.23\.attn_output\.weight$=q8_0
+^blk\.23\.attn_q\.weight$=q6_k
+^blk\.23\.attn_v\.weight$=q8_0
+^blk\.23\.ffn_down\.weight$=q6_k
+^blk\.23\.ffn_gate\.weight$=q6_k
+^blk\.23\.ffn_up\.weight$=q6_k
+^blk\.24\.attn_k\.weight$=q8_0
+^blk\.24\.attn_output\.weight$=q8_0
+^blk\.24\.attn_q\.weight$=q6_k
+^blk\.24\.attn_v\.weight$=q8_0
+^blk\.24\.ffn_down\.weight$=q6_k
+^blk\.24\.ffn_gate\.weight$=q6_k
+^blk\.24\.ffn_up\.weight$=q6_k
+^blk\.25\.attn_k\.weight$=q8_0
+^blk\.25\.attn_output\.weight$=q8_0
+^blk\.25\.attn_q\.weight$=q6_k
+^blk\.25\.attn_v\.weight$=q8_0
+^blk\.25\.ffn_down\.weight$=q6_k
+^blk\.25\.ffn_gate\.weight$=q6_k
+^blk\.25\.ffn_up\.weight$=q6_k
+^blk\.26\.attn_k\.weight$=q8_0
+^blk\.26\.attn_output\.weight$=q8_0
+^blk\.26\.attn_q\.weight$=q6_k
+^blk\.26\.attn_v\.weight$=q8_0
+^blk\.26\.ffn_down\.weight$=q6_k
+^blk\.26\.ffn_gate\.weight$=q6_k
+^blk\.26\.ffn_up\.weight$=q6_k
+^blk\.27\.attn_k\.weight$=q8_0
+^blk\.27\.attn_output\.weight$=q6_k
+^blk\.27\.attn_q\.weight$=q6_k
+^blk\.27\.attn_v\.weight$=q8_0
+^blk\.27\.ffn_down\.weight$=q6_k
+^blk\.27\.ffn_gate\.weight$=q6_k
+^blk\.27\.ffn_up\.weight$=q6_k
+^blk\.28\.attn_k\.weight$=q8_0
+^blk\.28\.attn_output\.weight$=q6_k
+^blk\.28\.attn_q\.weight$=q6_k
+^blk\.28\.attn_v\.weight$=q8_0
+^blk\.28\.ffn_down\.weight$=q6_k
+^blk\.28\.ffn_gate\.weight$=q6_k
+^blk\.28\.ffn_up\.weight$=q6_k
+^blk\.29\.attn_k\.weight$=q8_0
+^blk\.29\.attn_output\.weight$=q6_k
+^blk\.29\.attn_q\.weight$=q6_k
+^blk\.29\.attn_v\.weight$=q8_0
+^blk\.29\.ffn_down\.weight$=q6_k
+^blk\.29\.ffn_gate\.weight$=q6_k
+^blk\.29\.ffn_up\.weight$=q6_k
+^blk\.30\.attn_k\.weight$=q8_0
+^blk\.30\.attn_output\.weight$=q6_k
+^blk\.30\.attn_q\.weight$=q6_k
+^blk\.30\.attn_v\.weight$=q8_0
+^blk\.30\.ffn_down\.weight$=q6_k
+^blk\.30\.ffn_gate\.weight$=q6_k
+^blk\.30\.ffn_up\.weight$=q6_k
+^blk\.31\.attn_k\.weight$=q8_0
+^blk\.31\.attn_output\.weight$=q6_k
+^blk\.31\.attn_q\.weight$=q6_k
+^blk\.31\.attn_v\.weight$=q8_0
+^blk\.31\.ffn_down\.weight$=q6_k
+^blk\.31\.ffn_gate\.weight$=q6_k
+^blk\.31\.ffn_up\.weight$=q6_k
+^blk\.32\.attn_k\.weight$=q8_0
+^blk\.32\.attn_output\.weight$=q6_k
+^blk\.32\.attn_q\.weight$=q6_k
+^blk\.32\.attn_v\.weight$=q8_0
+^blk\.32\.ffn_down\.weight$=q6_k
+^blk\.32\.ffn_gate\.weight$=q6_k
+^blk\.32\.ffn_up\.weight$=q6_k
+^blk\.33\.attn_k\.weight$=q8_0
+^blk\.33\.attn_output\.weight$=q6_k
+^blk\.33\.attn_q\.weight$=q6_k
+^blk\.33\.attn_v\.weight$=q8_0
+^blk\.33\.ffn_down\.weight$=q6_k
+^blk\.33\.ffn_gate\.weight$=q6_k
+^blk\.33\.ffn_up\.weight$=q6_k
+^blk\.34\.attn_k\.weight$=q8_0
+^blk\.34\.attn_output\.weight$=q6_k
+^blk\.34\.attn_q\.weight$=q6_k
+^blk\.34\.attn_v\.weight$=q8_0
+^blk\.34\.ffn_down\.weight$=q6_k
+^blk\.34\.ffn_gate\.weight$=q6_k
+^blk\.34\.ffn_up\.weight$=q6_k
+^blk\.35\.attn_k\.weight$=q8_0
+^blk\.35\.attn_output\.weight$=q6_k
+^blk\.35\.attn_q\.weight$=q6_k
+^blk\.35\.attn_v\.weight$=q8_0
+^blk\.35\.ffn_down\.weight$=q8_0
+^blk\.35\.ffn_gate\.weight$=q6_k
+^blk\.35\.ffn_up\.weight$=q8_0

layouts/AesCode-8B-Q6_K_L.layout.json

@@ -0,0 +1,84 @@
+{
+  "model": "/media/m2teamgroup/models/AesCode-8B-GGUF/AesCode-8B-bf16.gguf",
+  "imatrix": "/media/m2teamgroup/models/AesCode-8B-GGUF/AesCode-8B-imatrix.gguf",
+  "ftype": "Q6_K_L",
+  "base_type": "q6_k",
+  "family": "k",
+  "recipe": "ident",
+  "variant": "shr0.50",
+  "auto_crush": false,
+  "rung_mode": "share",
+  "tier": "L",
+  "rung_share": 0.5,
+  "crush_pass_bytes": null,
+  "embd_shares": {
+    "output.weight": {
+      "share": 0.073,
+      "type": "q6_k"
+    },
+    "token_embd.weight": {
+      "share": 0.073,
+      "type": "q6_k"
+    }
+  },
+  "arch": "dense",
+  "tied": false,
+  "n_layer": 36,
+  "body_rule": true,
+  "pins": {},
+  "shape_pins": {},
+  "fixed_pins": {},
+  "embeddings": {
+    "output.weight": "q6_k",
+    "token_embd.weight": "q6_k"
+  },
+  "bytes": {
+    "heuristic_body+embd": 6992977920,
+    "map_body+embd": 7450091520,
+    "ratio": 1.0654,
+    "note": "quantizable, unpinned tensors only; pinned/unquantized tensors are identical in both"
+  },
+  "body_base_share": 0.501,
+  "body_bpw": 7.405,
+  "body_bpw_heuristic": 6.878,
+  "total_params": 8190735360,
+  "file_bytes_pred": 7451324416,
+  "file_bpw_pred": 7.278,
+  "histogram": {
+    "attn_k": {
+      "q8_0": 36
+    },
+    "attn_output": {
+      "q8_0": 27,
+      "q6_k": 9
+    },
+    "attn_q": {
+      "q6_k": 36
+    },
+    "attn_v": {
+      "q8_0": 36
+    },
+    "ffn_down": {
+      "q8_0": 18,
+      "q6_k": 18
+    },
+    "ffn_gate": {
+      "q6_k": 27,
+      "q8_0": 9
+    },
+    "ffn_up": {
+      "q8_0": 18,
+      "q6_k": 18
+    }
+  },
+  "prior": "layout/prior.json",
+  "pin_rule": "floor",
+  "embd_cap": "heur",
+  "generator": "auto_quant_v2.layout",
+  "generator_key": "3bf8b43e20a68f0b",
+  "quant_type": "Q6_K_L",
+  "quantize_ftype": "Q6_K",
+  "llama_cpp_version": "b11490",
+  "unknown_kinds": {},
+  "commit": "5d550ef5561a30307f1d436db2c6835b75395e77"
+}

layouts/AesCode-8B-Q6_K_L.tensor-types.txt

@@ -0,0 +1,254 @@
+^output\.weight$=q6_k
+^token_embd\.weight$=q6_k
+^blk\.0\.attn_k\.weight$=q8_0
+^blk\.0\.attn_output\.weight$=q8_0
+^blk\.0\.attn_q\.weight$=q6_k
+^blk\.0\.attn_v\.weight$=q8_0
+^blk\.0\.ffn_down\.weight$=q8_0
+^blk\.0\.ffn_gate\.weight$=q6_k
+^blk\.0\.ffn_up\.weight$=q8_0
+^blk\.1\.attn_k\.weight$=q8_0
+^blk\.1\.attn_output\.weight$=q8_0
+^blk\.1\.attn_q\.weight$=q6_k
+^blk\.1\.attn_v\.weight$=q8_0
+^blk\.1\.ffn_down\.weight$=q8_0
+^blk\.1\.ffn_gate\.weight$=q8_0
+^blk\.1\.ffn_up\.weight$=q8_0
+^blk\.2\.attn_k\.weight$=q8_0
+^blk\.2\.attn_output\.weight$=q8_0
+^blk\.2\.attn_q\.weight$=q6_k
+^blk\.2\.attn_v\.weight$=q8_0
+^blk\.2\.ffn_down\.weight$=q8_0
+^blk\.2\.ffn_gate\.weight$=q8_0
+^blk\.2\.ffn_up\.weight$=q8_0
+^blk\.3\.attn_k\.weight$=q8_0
+^blk\.3\.attn_output\.weight$=q8_0
+^blk\.3\.attn_q\.weight$=q6_k
+^blk\.3\.attn_v\.weight$=q8_0
+^blk\.3\.ffn_down\.weight$=q8_0
+^blk\.3\.ffn_gate\.weight$=q8_0
+^blk\.3\.ffn_up\.weight$=q8_0
+^blk\.4\.attn_k\.weight$=q8_0
+^blk\.4\.attn_output\.weight$=q8_0
+^blk\.4\.attn_q\.weight$=q6_k
+^blk\.4\.attn_v\.weight$=q8_0
+^blk\.4\.ffn_down\.weight$=q8_0
+^blk\.4\.ffn_gate\.weight$=q8_0
+^blk\.4\.ffn_up\.weight$=q8_0
+^blk\.5\.attn_k\.weight$=q8_0
+^blk\.5\.attn_output\.weight$=q8_0
+^blk\.5\.attn_q\.weight$=q6_k
+^blk\.5\.attn_v\.weight$=q8_0
+^blk\.5\.ffn_down\.weight$=q8_0
+^blk\.5\.ffn_gate\.weight$=q8_0
+^blk\.5\.ffn_up\.weight$=q8_0
+^blk\.6\.attn_k\.weight$=q8_0
+^blk\.6\.attn_output\.weight$=q8_0
+^blk\.6\.attn_q\.weight$=q6_k
+^blk\.6\.attn_v\.weight$=q8_0
+^blk\.6\.ffn_down\.weight$=q8_0
+^blk\.6\.ffn_gate\.weight$=q8_0
+^blk\.6\.ffn_up\.weight$=q8_0
+^blk\.7\.attn_k\.weight$=q8_0
+^blk\.7\.attn_output\.weight$=q8_0
+^blk\.7\.attn_q\.weight$=q6_k
+^blk\.7\.attn_v\.weight$=q8_0
+^blk\.7\.ffn_down\.weight$=q8_0
+^blk\.7\.ffn_gate\.weight$=q8_0
+^blk\.7\.ffn_up\.weight$=q8_0
+^blk\.8\.attn_k\.weight$=q8_0
+^blk\.8\.attn_output\.weight$=q8_0
+^blk\.8\.attn_q\.weight$=q6_k
+^blk\.8\.attn_v\.weight$=q8_0
+^blk\.8\.ffn_down\.weight$=q8_0
+^blk\.8\.ffn_gate\.weight$=q8_0
+^blk\.8\.ffn_up\.weight$=q8_0
+^blk\.9\.attn_k\.weight$=q8_0
+^blk\.9\.attn_output\.weight$=q8_0
+^blk\.9\.attn_q\.weight$=q6_k
+^blk\.9\.attn_v\.weight$=q8_0
+^blk\.9\.ffn_down\.weight$=q6_k
+^blk\.9\.ffn_gate\.weight$=q6_k
+^blk\.9\.ffn_up\.weight$=q6_k
+^blk\.10\.attn_k\.weight$=q8_0
+^blk\.10\.attn_output\.weight$=q8_0
+^blk\.10\.attn_q\.weight$=q6_k
+^blk\.10\.attn_v\.weight$=q8_0
+^blk\.10\.ffn_down\.weight$=q6_k
+^blk\.10\.ffn_gate\.weight$=q6_k
+^blk\.10\.ffn_up\.weight$=q6_k
+^blk\.11\.attn_k\.weight$=q8_0
+^blk\.11\.attn_output\.weight$=q8_0
+^blk\.11\.attn_q\.weight$=q6_k
+^blk\.11\.attn_v\.weight$=q8_0
+^blk\.11\.ffn_down\.weight$=q6_k
+^blk\.11\.ffn_gate\.weight$=q6_k
+^blk\.11\.ffn_up\.weight$=q6_k
+^blk\.12\.attn_k\.weight$=q8_0
+^blk\.12\.attn_output\.weight$=q8_0
+^blk\.12\.attn_q\.weight$=q6_k
+^blk\.12\.attn_v\.weight$=q8_0
+^blk\.12\.ffn_down\.weight$=q6_k
+^blk\.12\.ffn_gate\.weight$=q6_k
+^blk\.12\.ffn_up\.weight$=q6_k
+^blk\.13\.attn_k\.weight$=q8_0
+^blk\.13\.attn_output\.weight$=q8_0
+^blk\.13\.attn_q\.weight$=q6_k
+^blk\.13\.attn_v\.weight$=q8_0
+^blk\.13\.ffn_down\.weight$=q6_k
+^blk\.13\.ffn_gate\.weight$=q6_k
+^blk\.13\.ffn_up\.weight$=q6_k
+^blk\.14\.attn_k\.weight$=q8_0
+^blk\.14\.attn_output\.weight$=q8_0
+^blk\.14\.attn_q\.weight$=q6_k
+^blk\.14\.attn_v\.weight$=q8_0
+^blk\.14\.ffn_down\.weight$=q6_k
+^blk\.14\.ffn_gate\.weight$=q6_k
+^blk\.14\.ffn_up\.weight$=q6_k
+^blk\.15\.attn_k\.weight$=q8_0
+^blk\.15\.attn_output\.weight$=q8_0
+^blk\.15\.attn_q\.weight$=q6_k
+^blk\.15\.attn_v\.weight$=q8_0
+^blk\.15\.ffn_down\.weight$=q6_k
+^blk\.15\.ffn_gate\.weight$=q6_k
+^blk\.15\.ffn_up\.weight$=q6_k
+^blk\.16\.attn_k\.weight$=q8_0
+^blk\.16\.attn_output\.weight$=q8_0
+^blk\.16\.attn_q\.weight$=q6_k
+^blk\.16\.attn_v\.weight$=q8_0
+^blk\.16\.ffn_down\.weight$=q6_k
+^blk\.16\.ffn_gate\.weight$=q6_k
+^blk\.16\.ffn_up\.weight$=q6_k
+^blk\.17\.attn_k\.weight$=q8_0
+^blk\.17\.attn_output\.weight$=q8_0
+^blk\.17\.attn_q\.weight$=q6_k
+^blk\.17\.attn_v\.weight$=q8_0
+^blk\.17\.ffn_down\.weight$=q6_k
+^blk\.17\.ffn_gate\.weight$=q6_k
+^blk\.17\.ffn_up\.weight$=q6_k
+^blk\.18\.attn_k\.weight$=q8_0
+^blk\.18\.attn_output\.weight$=q8_0
+^blk\.18\.attn_q\.weight$=q6_k
+^blk\.18\.attn_v\.weight$=q8_0
+^blk\.18\.ffn_down\.weight$=q6_k
+^blk\.18\.ffn_gate\.weight$=q6_k
+^blk\.18\.ffn_up\.weight$=q6_k
+^blk\.19\.attn_k\.weight$=q8_0
+^blk\.19\.attn_output\.weight$=q8_0
+^blk\.19\.attn_q\.weight$=q6_k
+^blk\.19\.attn_v\.weight$=q8_0
+^blk\.19\.ffn_down\.weight$=q6_k
+^blk\.19\.ffn_gate\.weight$=q6_k
+^blk\.19\.ffn_up\.weight$=q6_k
+^blk\.20\.attn_k\.weight$=q8_0
+^blk\.20\.attn_output\.weight$=q8_0
+^blk\.20\.attn_q\.weight$=q6_k
+^blk\.20\.attn_v\.weight$=q8_0
+^blk\.20\.ffn_down\.weight$=q6_k
+^blk\.20\.ffn_gate\.weight$=q6_k
+^blk\.20\.ffn_up\.weight$=q6_k
+^blk\.21\.attn_k\.weight$=q8_0
+^blk\.21\.attn_output\.weight$=q8_0
+^blk\.21\.attn_q\.weight$=q6_k
+^blk\.21\.attn_v\.weight$=q8_0
+^blk\.21\.ffn_down\.weight$=q6_k
+^blk\.21\.ffn_gate\.weight$=q6_k
+^blk\.21\.ffn_up\.weight$=q6_k
+^blk\.22\.attn_k\.weight$=q8_0
+^blk\.22\.attn_output\.weight$=q8_0
+^blk\.22\.attn_q\.weight$=q6_k
+^blk\.22\.attn_v\.weight$=q8_0
+^blk\.22\.ffn_down\.weight$=q6_k
+^blk\.22\.ffn_gate\.weight$=q6_k
+^blk\.22\.ffn_up\.weight$=q6_k
+^blk\.23\.attn_k\.weight$=q8_0
+^blk\.23\.attn_output\.weight$=q8_0
+^blk\.23\.attn_q\.weight$=q6_k
+^blk\.23\.attn_v\.weight$=q8_0
+^blk\.23\.ffn_down\.weight$=q6_k
+^blk\.23\.ffn_gate\.weight$=q6_k
+^blk\.23\.ffn_up\.weight$=q6_k
+^blk\.24\.attn_k\.weight$=q8_0
+^blk\.24\.attn_output\.weight$=q8_0
+^blk\.24\.attn_q\.weight$=q6_k
+^blk\.24\.attn_v\.weight$=q8_0
+^blk\.24\.ffn_down\.weight$=q6_k
+^blk\.24\.ffn_gate\.weight$=q6_k
+^blk\.24\.ffn_up\.weight$=q6_k
+^blk\.25\.attn_k\.weight$=q8_0
+^blk\.25\.attn_output\.weight$=q8_0
+^blk\.25\.attn_q\.weight$=q6_k
+^blk\.25\.attn_v\.weight$=q8_0
+^blk\.25\.ffn_down\.weight$=q6_k
+^blk\.25\.ffn_gate\.weight$=q6_k
+^blk\.25\.ffn_up\.weight$=q6_k
+^blk\.26\.attn_k\.weight$=q8_0
+^blk\.26\.attn_output\.weight$=q8_0
+^blk\.26\.attn_q\.weight$=q6_k
+^blk\.26\.attn_v\.weight$=q8_0
+^blk\.26\.ffn_down\.weight$=q6_k
+^blk\.26\.ffn_gate\.weight$=q6_k
+^blk\.26\.ffn_up\.weight$=q6_k
+^blk\.27\.attn_k\.weight$=q8_0
+^blk\.27\.attn_output\.weight$=q6_k
+^blk\.27\.attn_q\.weight$=q6_k
+^blk\.27\.attn_v\.weight$=q8_0
+^blk\.27\.ffn_down\.weight$=q8_0
+^blk\.27\.ffn_gate\.weight$=q6_k
+^blk\.27\.ffn_up\.weight$=q8_0
+^blk\.28\.attn_k\.weight$=q8_0
+^blk\.28\.attn_output\.weight$=q6_k
+^blk\.28\.attn_q\.weight$=q6_k
+^blk\.28\.attn_v\.weight$=q8_0
+^blk\.28\.ffn_down\.weight$=q8_0
+^blk\.28\.ffn_gate\.weight$=q6_k
+^blk\.28\.ffn_up\.weight$=q8_0
+^blk\.29\.attn_k\.weight$=q8_0
+^blk\.29\.attn_output\.weight$=q6_k
+^blk\.29\.attn_q\.weight$=q6_k
+^blk\.29\.attn_v\.weight$=q8_0
+^blk\.29\.ffn_down\.weight$=q8_0
+^blk\.29\.ffn_gate\.weight$=q6_k
+^blk\.29\.ffn_up\.weight$=q8_0
+^blk\.30\.attn_k\.weight$=q8_0
+^blk\.30\.attn_output\.weight$=q6_k
+^blk\.30\.attn_q\.weight$=q6_k
+^blk\.30\.attn_v\.weight$=q8_0
+^blk\.30\.ffn_down\.weight$=q8_0
+^blk\.30\.ffn_gate\.weight$=q6_k
+^blk\.30\.ffn_up\.weight$=q8_0
+^blk\.31\.attn_k\.weight$=q8_0
+^blk\.31\.attn_output\.weight$=q6_k
+^blk\.31\.attn_q\.weight$=q6_k
+^blk\.31\.attn_v\.weight$=q8_0
+^blk\.31\.ffn_down\.weight$=q8_0
+^blk\.31\.ffn_gate\.weight$=q6_k
+^blk\.31\.ffn_up\.weight$=q8_0
+^blk\.32\.attn_k\.weight$=q8_0
+^blk\.32\.attn_output\.weight$=q6_k
+^blk\.32\.attn_q\.weight$=q6_k
+^blk\.32\.attn_v\.weight$=q8_0
+^blk\.32\.ffn_down\.weight$=q8_0
+^blk\.32\.ffn_gate\.weight$=q6_k
+^blk\.32\.ffn_up\.weight$=q8_0
+^blk\.33\.attn_k\.weight$=q8_0
+^blk\.33\.attn_output\.weight$=q6_k
+^blk\.33\.attn_q\.weight$=q6_k
+^blk\.33\.attn_v\.weight$=q8_0
+^blk\.33\.ffn_down\.weight$=q8_0
+^blk\.33\.ffn_gate\.weight$=q6_k
+^blk\.33\.ffn_up\.weight$=q8_0
+^blk\.34\.attn_k\.weight$=q8_0
+^blk\.34\.attn_output\.weight$=q6_k
+^blk\.34\.attn_q\.weight$=q6_k
+^blk\.34\.attn_v\.weight$=q8_0
+^blk\.34\.ffn_down\.weight$=q8_0
+^blk\.34\.ffn_gate\.weight$=q6_k
+^blk\.34\.ffn_up\.weight$=q8_0
+^blk\.35\.attn_k\.weight$=q8_0
+^blk\.35\.attn_output\.weight$=q6_k
+^blk\.35\.attn_q\.weight$=q6_k
+^blk\.35\.attn_v\.weight$=q8_0
+^blk\.35\.ffn_down\.weight$=q8_0
+^blk\.35\.ffn_gate\.weight$=q8_0
+^blk\.35\.ffn_up\.weight$=q8_0

layouts/AesCode-8B-Q6_K_S.layout.json

@@ -0,0 +1,84 @@
+{
+  "model": "/media/m2teamgroup/models/AesCode-8B-GGUF/AesCode-8B-bf16.gguf",
+  "imatrix": "/media/m2teamgroup/models/AesCode-8B-GGUF/AesCode-8B-imatrix.gguf",
+  "ftype": "Q6_K_S",
+  "base_type": "q6_k",
+  "family": "k",
+  "recipe": "ident",
+  "variant": "shr0.90",
+  "auto_crush": false,
+  "rung_mode": "share",
+  "tier": "S",
+  "rung_share": 0.9,
+  "crush_pass_bytes": null,
+  "embd_shares": {
+    "output.weight": {
+      "share": 0.073,
+      "type": "q6_k"
+    },
+    "token_embd.weight": {
+      "share": 0.073,
+      "type": "q6_k"
+    }
+  },
+  "arch": "dense",
+  "tied": false,
+  "n_layer": 36,
+  "body_rule": true,
+  "pins": {},
+  "shape_pins": {},
+  "fixed_pins": {},
+  "embeddings": {
+    "output.weight": "q6_k",
+    "token_embd.weight": "q6_k"
+  },
+  "bytes": {
+    "heuristic_body+embd": 6992977920,
+    "map_body+embd": 6850764800,
+    "ratio": 0.9797,
+    "note": "quantizable, unpinned tensors only; pinned/unquantized tensors are identical in both"
+  },
+  "body_base_share": 0.901,
+  "body_bpw": 6.715,
+  "body_bpw_heuristic": 6.878,
+  "total_params": 8190735360,
+  "file_bytes_pred": 6851997696,
+  "file_bpw_pred": 6.692,
+  "histogram": {
+    "attn_k": {
+      "q8_0": 30,
+      "q6_k": 6
+    },
+    "attn_output": {
+      "q6_k": 26,
+      "q8_0": 10
+    },
+    "attn_q": {
+      "q6_k": 36
+    },
+    "attn_v": {
+      "q8_0": 36
+    },
+    "ffn_down": {
+      "q6_k": 35,
+      "q8_0": 1
+    },
+    "ffn_gate": {
+      "q6_k": 36
+    },
+    "ffn_up": {
+      "q6_k": 35,
+      "q8_0": 1
+    }
+  },
+  "prior": "layout/prior.json",
+  "pin_rule": "floor",
+  "embd_cap": "heur",
+  "generator": "auto_quant_v2.layout",
+  "generator_key": "3bf8b43e20a68f0b",
+  "quant_type": "Q6_K_S",
+  "quantize_ftype": "Q6_K",
+  "llama_cpp_version": "b11490",
+  "unknown_kinds": {},
+  "commit": "5d550ef5561a30307f1d436db2c6835b75395e77"
+}

layouts/AesCode-8B-Q6_K_S.tensor-types.txt

@@ -0,0 +1,254 @@
+^output\.weight$=q6_k
+^token_embd\.weight$=q6_k
+^blk\.0\.attn_k\.weight$=q8_0
+^blk\.0\.attn_output\.weight$=q6_k
+^blk\.0\.attn_q\.weight$=q6_k
+^blk\.0\.attn_v\.weight$=q8_0
+^blk\.0\.ffn_down\.weight$=q6_k
+^blk\.0\.ffn_gate\.weight$=q6_k
+^blk\.0\.ffn_up\.weight$=q6_k
+^blk\.1\.attn_k\.weight$=q8_0
+^blk\.1\.attn_output\.weight$=q8_0
+^blk\.1\.attn_q\.weight$=q6_k
+^blk\.1\.attn_v\.weight$=q8_0
+^blk\.1\.ffn_down\.weight$=q6_k
+^blk\.1\.ffn_gate\.weight$=q6_k
+^blk\.1\.ffn_up\.weight$=q6_k
+^blk\.2\.attn_k\.weight$=q8_0
+^blk\.2\.attn_output\.weight$=q8_0
+^blk\.2\.attn_q\.weight$=q6_k
+^blk\.2\.attn_v\.weight$=q8_0
+^blk\.2\.ffn_down\.weight$=q6_k
+^blk\.2\.ffn_gate\.weight$=q6_k
+^blk\.2\.ffn_up\.weight$=q6_k
+^blk\.3\.attn_k\.weight$=q8_0
+^blk\.3\.attn_output\.weight$=q8_0
+^blk\.3\.attn_q\.weight$=q6_k
+^blk\.3\.attn_v\.weight$=q8_0
+^blk\.3\.ffn_down\.weight$=q6_k
+^blk\.3\.ffn_gate\.weight$=q6_k
+^blk\.3\.ffn_up\.weight$=q6_k
+^blk\.4\.attn_k\.weight$=q8_0
+^blk\.4\.attn_output\.weight$=q8_0
+^blk\.4\.attn_q\.weight$=q6_k
+^blk\.4\.attn_v\.weight$=q8_0
+^blk\.4\.ffn_down\.weight$=q6_k
+^blk\.4\.ffn_gate\.weight$=q6_k
+^blk\.4\.ffn_up\.weight$=q6_k
+^blk\.5\.attn_k\.weight$=q8_0
+^blk\.5\.attn_output\.weight$=q8_0
+^blk\.5\.attn_q\.weight$=q6_k
+^blk\.5\.attn_v\.weight$=q8_0
+^blk\.5\.ffn_down\.weight$=q6_k
+^blk\.5\.ffn_gate\.weight$=q6_k
+^blk\.5\.ffn_up\.weight$=q6_k
+^blk\.6\.attn_k\.weight$=q8_0
+^blk\.6\.attn_output\.weight$=q8_0
+^blk\.6\.attn_q\.weight$=q6_k
+^blk\.6\.attn_v\.weight$=q8_0
+^blk\.6\.ffn_down\.weight$=q6_k
+^blk\.6\.ffn_gate\.weight$=q6_k
+^blk\.6\.ffn_up\.weight$=q6_k
+^blk\.7\.attn_k\.weight$=q8_0
+^blk\.7\.attn_output\.weight$=q8_0
+^blk\.7\.attn_q\.weight$=q6_k
+^blk\.7\.attn_v\.weight$=q8_0
+^blk\.7\.ffn_down\.weight$=q6_k
+^blk\.7\.ffn_gate\.weight$=q6_k
+^blk\.7\.ffn_up\.weight$=q6_k
+^blk\.8\.attn_k\.weight$=q8_0
+^blk\.8\.attn_output\.weight$=q8_0
+^blk\.8\.attn_q\.weight$=q6_k
+^blk\.8\.attn_v\.weight$=q8_0
+^blk\.8\.ffn_down\.weight$=q6_k
+^blk\.8\.ffn_gate\.weight$=q6_k
+^blk\.8\.ffn_up\.weight$=q6_k
+^blk\.9\.attn_k\.weight$=q8_0
+^blk\.9\.attn_output\.weight$=q8_0
+^blk\.9\.attn_q\.weight$=q6_k
+^blk\.9\.attn_v\.weight$=q8_0
+^blk\.9\.ffn_down\.weight$=q6_k
+^blk\.9\.ffn_gate\.weight$=q6_k
+^blk\.9\.ffn_up\.weight$=q6_k
+^blk\.10\.attn_k\.weight$=q8_0
+^blk\.10\.attn_output\.weight$=q8_0
+^blk\.10\.attn_q\.weight$=q6_k
+^blk\.10\.attn_v\.weight$=q8_0
+^blk\.10\.ffn_down\.weight$=q6_k
+^blk\.10\.ffn_gate\.weight$=q6_k
+^blk\.10\.ffn_up\.weight$=q6_k
+^blk\.11\.attn_k\.weight$=q8_0
+^blk\.11\.attn_output\.weight$=q6_k
+^blk\.11\.attn_q\.weight$=q6_k
+^blk\.11\.attn_v\.weight$=q8_0
+^blk\.11\.ffn_down\.weight$=q6_k
+^blk\.11\.ffn_gate\.weight$=q6_k
+^blk\.11\.ffn_up\.weight$=q6_k
+^blk\.12\.attn_k\.weight$=q8_0
+^blk\.12\.attn_output\.weight$=q6_k
+^blk\.12\.attn_q\.weight$=q6_k
+^blk\.12\.attn_v\.weight$=q8_0
+^blk\.12\.ffn_down\.weight$=q6_k
+^blk\.12\.ffn_gate\.weight$=q6_k
+^blk\.12\.ffn_up\.weight$=q6_k
+^blk\.13\.attn_k\.weight$=q8_0
+^blk\.13\.attn_output\.weight$=q6_k
+^blk\.13\.attn_q\.weight$=q6_k
+^blk\.13\.attn_v\.weight$=q8_0
+^blk\.13\.ffn_down\.weight$=q6_k
+^blk\.13\.ffn_gate\.weight$=q6_k
+^blk\.13\.ffn_up\.weight$=q6_k
+^blk\.14\.attn_k\.weight$=q8_0
+^blk\.14\.attn_output\.weight$=q6_k
+^blk\.14\.attn_q\.weight$=q6_k
+^blk\.14\.attn_v\.weight$=q8_0
+^blk\.14\.ffn_down\.weight$=q6_k
+^blk\.14\.ffn_gate\.weight$=q6_k
+^blk\.14\.ffn_up\.weight$=q6_k
+^blk\.15\.attn_k\.weight$=q8_0
+^blk\.15\.attn_output\.weight$=q6_k
+^blk\.15\.attn_q\.weight$=q6_k
+^blk\.15\.attn_v\.weight$=q8_0
+^blk\.15\.ffn_down\.weight$=q6_k
+^blk\.15\.ffn_gate\.weight$=q6_k
+^blk\.15\.ffn_up\.weight$=q6_k
+^blk\.16\.attn_k\.weight$=q8_0
+^blk\.16\.attn_output\.weight$=q6_k
+^blk\.16\.attn_q\.weight$=q6_k
+^blk\.16\.attn_v\.weight$=q8_0
+^blk\.16\.ffn_down\.weight$=q6_k
+^blk\.16\.ffn_gate\.weight$=q6_k
+^blk\.16\.ffn_up\.weight$=q6_k
+^blk\.17\.attn_k\.weight$=q8_0
+^blk\.17\.attn_output\.weight$=q6_k
+^blk\.17\.attn_q\.weight$=q6_k
+^blk\.17\.attn_v\.weight$=q8_0
+^blk\.17\.ffn_down\.weight$=q6_k
+^blk\.17\.ffn_gate\.weight$=q6_k
+^blk\.17\.ffn_up\.weight$=q6_k
+^blk\.18\.attn_k\.weight$=q8_0
+^blk\.18\.attn_output\.weight$=q6_k
+^blk\.18\.attn_q\.weight$=q6_k
+^blk\.18\.attn_v\.weight$=q8_0
+^blk\.18\.ffn_down\.weight$=q6_k
+^blk\.18\.ffn_gate\.weight$=q6_k
+^blk\.18\.ffn_up\.weight$=q6_k
+^blk\.19\.attn_k\.weight$=q8_0
+^blk\.19\.attn_output\.weight$=q6_k
+^blk\.19\.attn_q\.weight$=q6_k
+^blk\.19\.attn_v\.weight$=q8_0
+^blk\.19\.ffn_down\.weight$=q6_k
+^blk\.19\.ffn_gate\.weight$=q6_k
+^blk\.19\.ffn_up\.weight$=q6_k
+^blk\.20\.attn_k\.weight$=q8_0
+^blk\.20\.attn_output\.weight$=q6_k
+^blk\.20\.attn_q\.weight$=q6_k
+^blk\.20\.attn_v\.weight$=q8_0
+^blk\.20\.ffn_down\.weight$=q6_k
+^blk\.20\.ffn_gate\.weight$=q6_k
+^blk\.20\.ffn_up\.weight$=q6_k
+^blk\.21\.attn_k\.weight$=q8_0
+^blk\.21\.attn_output\.weight$=q6_k
+^blk\.21\.attn_q\.weight$=q6_k
+^blk\.21\.attn_v\.weight$=q8_0
+^blk\.21\.ffn_down\.weight$=q6_k
+^blk\.21\.ffn_gate\.weight$=q6_k
+^blk\.21\.ffn_up\.weight$=q6_k
+^blk\.22\.attn_k\.weight$=q8_0
+^blk\.22\.attn_output\.weight$=q6_k
+^blk\.22\.attn_q\.weight$=q6_k
+^blk\.22\.attn_v\.weight$=q8_0
+^blk\.22\.ffn_down\.weight$=q6_k
+^blk\.22\.ffn_gate\.weight$=q6_k
+^blk\.22\.ffn_up\.weight$=q6_k
+^blk\.23\.attn_k\.weight$=q8_0
+^blk\.23\.attn_output\.weight$=q6_k
+^blk\.23\.attn_q\.weight$=q6_k
+^blk\.23\.attn_v\.weight$=q8_0
+^blk\.23\.ffn_down\.weight$=q6_k
+^blk\.23\.ffn_gate\.weight$=q6_k
+^blk\.23\.ffn_up\.weight$=q6_k
+^blk\.24\.attn_k\.weight$=q8_0
+^blk\.24\.attn_output\.weight$=q6_k
+^blk\.24\.attn_q\.weight$=q6_k
+^blk\.24\.attn_v\.weight$=q8_0
+^blk\.24\.ffn_down\.weight$=q6_k
+^blk\.24\.ffn_gate\.weight$=q6_k
+^blk\.24\.ffn_up\.weight$=q6_k
+^blk\.25\.attn_k\.weight$=q8_0
+^blk\.25\.attn_output\.weight$=q6_k
+^blk\.25\.attn_q\.weight$=q6_k
+^blk\.25\.attn_v\.weight$=q8_0
+^blk\.25\.ffn_down\.weight$=q6_k
+^blk\.25\.ffn_gate\.weight$=q6_k
+^blk\.25\.ffn_up\.weight$=q6_k
+^blk\.26\.attn_k\.weight$=q8_0
+^blk\.26\.attn_output\.weight$=q6_k
+^blk\.26\.attn_q\.weight$=q6_k
+^blk\.26\.attn_v\.weight$=q8_0
+^blk\.26\.ffn_down\.weight$=q6_k
+^blk\.26\.ffn_gate\.weight$=q6_k
+^blk\.26\.ffn_up\.weight$=q6_k
+^blk\.27\.attn_k\.weight$=q8_0
+^blk\.27\.attn_output\.weight$=q6_k
+^blk\.27\.attn_q\.weight$=q6_k
+^blk\.27\.attn_v\.weight$=q8_0
+^blk\.27\.ffn_down\.weight$=q6_k
+^blk\.27\.ffn_gate\.weight$=q6_k
+^blk\.27\.ffn_up\.weight$=q6_k
+^blk\.28\.attn_k\.weight$=q8_0
+^blk\.28\.attn_output\.weight$=q6_k
+^blk\.28\.attn_q\.weight$=q6_k
+^blk\.28\.attn_v\.weight$=q8_0
+^blk\.28\.ffn_down\.weight$=q6_k
+^blk\.28\.ffn_gate\.weight$=q6_k
+^blk\.28\.ffn_up\.weight$=q6_k
+^blk\.29\.attn_k\.weight$=q8_0
+^blk\.29\.attn_output\.weight$=q6_k
+^blk\.29\.attn_q\.weight$=q6_k
+^blk\.29\.attn_v\.weight$=q8_0
+^blk\.29\.ffn_down\.weight$=q6_k
+^blk\.29\.ffn_gate\.weight$=q6_k
+^blk\.29\.ffn_up\.weight$=q6_k
+^blk\.30\.attn_k\.weight$=q6_k
+^blk\.30\.attn_output\.weight$=q6_k
+^blk\.30\.attn_q\.weight$=q6_k
+^blk\.30\.attn_v\.weight$=q8_0
+^blk\.30\.ffn_down\.weight$=q6_k
+^blk\.30\.ffn_gate\.weight$=q6_k
+^blk\.30\.ffn_up\.weight$=q6_k
+^blk\.31\.attn_k\.weight$=q6_k
+^blk\.31\.attn_output\.weight$=q6_k
+^blk\.31\.attn_q\.weight$=q6_k
+^blk\.31\.attn_v\.weight$=q8_0
+^blk\.31\.ffn_down\.weight$=q6_k
+^blk\.31\.ffn_gate\.weight$=q6_k
+^blk\.31\.ffn_up\.weight$=q6_k
+^blk\.32\.attn_k\.weight$=q6_k
+^blk\.32\.attn_output\.weight$=q6_k
+^blk\.32\.attn_q\.weight$=q6_k
+^blk\.32\.attn_v\.weight$=q8_0
+^blk\.32\.ffn_down\.weight$=q6_k
+^blk\.32\.ffn_gate\.weight$=q6_k
+^blk\.32\.ffn_up\.weight$=q6_k
+^blk\.33\.attn_k\.weight$=q6_k
+^blk\.33\.attn_output\.weight$=q6_k
+^blk\.33\.attn_q\.weight$=q6_k
+^blk\.33\.attn_v\.weight$=q8_0
+^blk\.33\.ffn_down\.weight$=q6_k
+^blk\.33\.ffn_gate\.weight$=q6_k
+^blk\.33\.ffn_up\.weight$=q6_k
+^blk\.34\.attn_k\.weight$=q6_k
+^blk\.34\.attn_output\.weight$=q6_k
+^blk\.34\.attn_q\.weight$=q6_k
+^blk\.34\.attn_v\.weight$=q8_0
+^blk\.34\.ffn_down\.weight$=q6_k
+^blk\.34\.ffn_gate\.weight$=q6_k
+^blk\.34\.ffn_up\.weight$=q6_k
+^blk\.35\.attn_k\.weight$=q6_k
+^blk\.35\.attn_output\.weight$=q6_k
+^blk\.35\.attn_q\.weight$=q6_k
+^blk\.35\.attn_v\.weight$=q8_0
+^blk\.35\.ffn_down\.weight$=q8_0
+^blk\.35\.ffn_gate\.weight$=q6_k
+^blk\.35\.ffn_up\.weight$=q8_0