convert : allow partial update to the chkhsh pre-tokenizer list (#13847)

* convert : allow partial update to the chkhsh pre-tokenizer list

* code style

* update tokenizer out

* rm inp/out files for models not having gguf

* fixed hash for glm

* skip nomic-bert-moe test

* Update convert_hf_to_gguf_update.py

* fix minerva-7b hash

* rm redundant import

convert_hf_to_gguf.py

@@ -674,12 +674,12 @@ class TextModel(ModelBase):
         if chkhsh == "8aeee3860c56296a157a1fe2fad249ec40aa59b1bb5709f4ade11c4e6fe652ed":
             # ref: https://huggingface.co/tiiuae/falcon-7b
             res = "falcon"
-        if chkhsh == "9d032fcbd5501f4a38150912590928bfb36091efb5df11b8e2124b0390e3fb1e":
-            # ref: https://huggingface.co/tiiuae/Falcon3-7B-Base
-            res = "falcon3"
         if chkhsh == "0876d13b50744004aa9aeae05e7b0647eac9d801b5ba4668afc01e709c15e19f":
             # ref: https://huggingface.co/BAAI/bge-small-en-v1.5
             res = "bert-bge"
+        if chkhsh == "9d032fcbd5501f4a38150912590928bfb36091efb5df11b8e2124b0390e3fb1e":
+            # ref: https://huggingface.co/tiiuae/Falcon3-7B-Base
+            res = "falcon3"
         if chkhsh == "8e62295832751ca1e8f92f2226f403dea30dc5165e448b5bfa05af5340c64ec7":
             # ref: https://huggingface.co/BAAI/bge-large-zh-v1.5
             res = "bert-bge-large"
@@ -731,9 +731,6 @@ class TextModel(ModelBase):
         if chkhsh == "7967bfa498ade6b757b064f31e964dddbb80f8f9a4d68d4ba7998fcf281c531a":
             # ref: https://huggingface.co/jinaai/jina-embeddings-v2-base-code
             res = "jina-v2-code"
-        if chkhsh == "b6e8e1518dc4305be2fe39c313ed643381c4da5db34a98f6a04c093f8afbe99b" or chkhsh == "81d72c7348a9f0ebe86f23298d37debe0a5e71149e29bd283904c02262b27516":
-            # ref: https://huggingface.co/THUDM/glm-4-9b-chat
-            res = "chatglm-bpe"
         if chkhsh == "7fc505bd3104ca1083b150b17d088b59534ede9bde81f0dd2090967d7fe52cee":
             # ref: https://huggingface.co/LumiOpen/Viking-7B
             res = "viking"
@@ -764,9 +761,6 @@ class TextModel(ModelBase):
         if chkhsh == "60824e3c0d9401f89943cbb2fff727f0e2d4c545ba4df2d6e4f09a6db0f5b450":
             # ref: https://huggingface.co/facebook/chameleon-7b
             res = "chameleon"
-        if chkhsh == "1431a23e583c97432bc230bff598d103ddb5a1f89960c8f1d1051aaa944d0b35":
-            # ref: https://huggingface.co/sapienzanlp/Minerva-7B-base-v1.0
-            res = "minerva-7b"
         if chkhsh == "8b5a93ed704057481f240da0be7e7dca721d7f8f4755263b6807227a2cbeae65":
             # ref: https://huggingface.co/sentence-transformers/stsb-roberta-base
             res = "roberta-bpe"
@@ -797,15 +791,24 @@ class TextModel(ModelBase):
         if chkhsh == "d353350c764d8c3b39c763113960e4fb4919bea5fbf208a0e3b22e8469dc7406":
             # ref: https://huggingface.co/meta-llama/Llama-4-Scout-17B-16E-Instruct
             res = "llama4"
-        if chkhsh == "a1336059768a55c99a734006ffb02203cd450fed003e9a71886c88acf24fdbc2":
-            # ref: https://huggingface.co/THUDM/glm-4-9b-hf
-            res = "glm4"
         if chkhsh == "0e9433cbbb161f89e264eb32e8e64bfe69e834973ffca5d41d3948a604a3e2a3":
             # ref: https://huggingface.co/mistral-community/pixtral-12b
             res = "pixtral"
         if chkhsh == "d5f1dd6f980fec569fb218a81a7658ac45fc56b38c5a0adeb1c232fbe04ef5ec":
             # ref: https://huggingface.co/ByteDance-Seed/Seed-Coder-8B-Base
             res = "seed-coder"
+        if chkhsh == "b6e8e1518dc4305be2fe39c313ed643381c4da5db34a98f6a04c093f8afbe99b":
+            # ref: https://huggingface.co/THUDM/glm-4-9b-chat
+            res = "chatglm-bpe"
+        if chkhsh == "81d72c7348a9f0ebe86f23298d37debe0a5e71149e29bd283904c02262b27516":
+            # ref: https://huggingface.co/THUDM/glm-4-9b-chat
+            res = "chatglm-bpe"
+        if chkhsh == "a1336059768a55c99a734006ffb02203cd450fed003e9a71886c88acf24fdbc2":
+            # ref: https://huggingface.co/THUDM/glm-4-9b-hf
+            res = "glm4"
+        if chkhsh == "1431a23e583c97432bc230bff598d103ddb5a1f89960c8f1d1051aaa944d0b35":
+            # ref: https://huggingface.co/sapienzanlp/Minerva-7B-base-v1.0
+            res = "minerva-7b"
 
         if res is None:
             logger.warning("\n")

convert_hf_to_gguf_update.py

@@ -1,28 +1,6 @@
 #!/usr/bin/env python3
 # -*- coding: utf-8 -*-
 
-# This script downloads the tokenizer models of the specified models from Huggingface and
-# generates the get_vocab_base_pre() function for convert_hf_to_gguf.py
-#
-# This is necessary in order to analyze the type of pre-tokenizer used by the model and
-# provide the necessary information to llama.cpp via the GGUF header in order to implement
-# the same pre-tokenizer.
-#
-# ref: https://github.com/ggml-org/llama.cpp/pull/6920
-#
-# Instructions:
-#
-# - Add a new model to the "models" list
-# - Run the script with your huggingface token:
-#
-#   python3 convert_hf_to_gguf_update.py <huggingface_token>
-#
-# - The convert_hf_to_gguf.py script will have had its get_vocab_base_pre() function updated
-# - Update llama.cpp with the new pre-tokenizer if necessary
-#
-# TODO: generate tokenizer tests for llama.cpp
-#
-
 import logging
 import os
 import pathlib
@@ -32,6 +10,7 @@ import requests
 import sys
 import json
 import shutil
+import argparse
 
 from hashlib import sha256
 from enum import IntEnum, auto
@@ -41,6 +20,11 @@ logging.basicConfig(level=logging.DEBUG)
 logger = logging.getLogger("convert_hf_to_gguf_update")
 sess = requests.Session()
 
+convert_py_pth = pathlib.Path("convert_hf_to_gguf.py")
+convert_py = convert_py_pth.read_text(encoding="utf-8")
+hf_token_pth = pathlib.Path.home() / ".cache" / "huggingface" / "token"
+hf_token = hf_token_pth.read_text(encoding="utf-8").strip() if hf_token_pth.exists() else None
+
 
 class TOKENIZER_TYPE(IntEnum):
     SPM = auto()
@@ -49,20 +33,49 @@ class TOKENIZER_TYPE(IntEnum):
     UGM = auto()
 
 
+DOC_STRING = """
+This script downloads the tokenizer models of the specified models from Huggingface and
+generates the get_vocab_base_pre() function for convert_hf_to_gguf.py
+
+/!\\ It is intended to be used by contributors and is not meant to be run by end users
+
+This is necessary in order to analyze the type of pre-tokenizer used by the model and
+provide the necessary information to llama.cpp via the GGUF header in order to implement
+the same pre-tokenizer.
+
+ref: https://github.com/ggml-org/llama.cpp/pull/6920
+
+Instructions:
+
+- Add a new model to the "models" list
+- Run the script with your huggingface token
+    By default, token will be read from ~/.cache/huggingface/token
+- The convert_hf_to_gguf.py script will have had its get_vocab_base_pre() function updated
+- Update llama.cpp with the new pre-tokenizer if necessary
+"""
+# TODO: generate tokenizer tests for llama.cpp
+
+parser = argparse.ArgumentParser(description=DOC_STRING, formatter_class=argparse.RawTextHelpFormatter)
+parser.add_argument(
+    "--full", action="store_true",
+    help="download full list of models - make sure you have access to all of them",
+)
+parser.add_argument(
+    "hf_token",
+    help="optional HF token",
+    nargs="?",
+)
+args = parser.parse_args()
+hf_token = args.hf_token if args.hf_token is not None else hf_token
+
+if hf_token is None:
+    logger.error("HF token is required. Please provide it as an argument or set it in ~/.cache/huggingface/token")
+    sys.exit(1)
+
 # TODO: this string has to exercise as much pre-tokenizer functionality as possible
 #       will be updated with time - contributions welcome
 CHK_TXT = '\n \n\n \n\n\n \t \t\t \t\n  \n   \n    \n     \n🚀 (normal) 😶‍🌫️ (multiple emojis concatenated) ✅ 🦙🦙 3 33 333 3333 33333 333333 3333333 33333333 3.3 3..3 3...3 កាន់តែពិសេសអាច😁 ?我想在apple工作1314151天～ ------======= нещо на Български \'\'\'\'\'\'```````\"\"\"\"......!!!!!!?????? I\'ve been \'told he\'s there, \'RE you sure? \'M not sure I\'ll make it, \'D you like some tea? We\'Ve a\'lL'
 
-if len(sys.argv) == 2:
-    token = sys.argv[1]
-    if not token.startswith("hf_"):
-        logger.info("Huggingface token seems invalid")
-        logger.info("Usage: python convert_hf_to_gguf_update.py <huggingface_token>")
-        sys.exit(1)
-else:
-    logger.info("Usage: python convert_hf_to_gguf_update.py <huggingface_token>")
-    sys.exit(1)
-
 # TODO: add models here, base models preferred
 models = [
     {"name": "llama-spm",        "tokt": TOKENIZER_TYPE.SPM, "repo": "https://huggingface.co/meta-llama/Llama-2-7b-hf", },
@@ -103,7 +116,6 @@ models = [
     {"name": "exaone",           "tokt": TOKENIZER_TYPE.BPE, "repo": "https://huggingface.co/LGAI-EXAONE/EXAONE-3.0-7.8B-Instruct", },
     {"name": "phi-2",            "tokt": TOKENIZER_TYPE.BPE, "repo": "https://huggingface.co/microsoft/phi-2", },
     {"name": "chameleon",        "tokt": TOKENIZER_TYPE.BPE, "repo": "https://huggingface.co/facebook/chameleon-7b", },
-    {"name": "minerva-7b",       "tokt": TOKENIZER_TYPE.BPE, "repo": "https://huggingface.co/sapienzanlp/Minerva-7B-base-v1.0", },
     {"name": "roberta-bpe",      "tokt": TOKENIZER_TYPE.BPE, "repo": "https://huggingface.co/sentence-transformers/stsb-roberta-base"},
     {"name": "gigachat",         "tokt": TOKENIZER_TYPE.BPE, "repo": "https://huggingface.co/ai-sage/GigaChat-20B-A3B-instruct"},
     {"name": "megrez",           "tokt": TOKENIZER_TYPE.BPE, "repo": "https://huggingface.co/Infinigence/Megrez-3B-Instruct"},
@@ -114,11 +126,19 @@ models = [
     {"name": "trillion",         "tokt": TOKENIZER_TYPE.BPE, "repo": "https://huggingface.co/trillionlabs/Trillion-7B-preview", },
     {"name": "bailingmoe",       "tokt": TOKENIZER_TYPE.BPE, "repo": "https://huggingface.co/inclusionAI/Ling-lite", },
     {"name": "llama4",           "tokt": TOKENIZER_TYPE.BPE, "repo": "https://huggingface.co/meta-llama/Llama-4-Scout-17B-16E-Instruct", },
-    {"name": "glm4",             "tokt": TOKENIZER_TYPE.BPE, "repo": "https://huggingface.co/THUDM/glm-4-9b-hf", },
     {"name": "pixtral",          "tokt": TOKENIZER_TYPE.BPE, "repo": "https://huggingface.co/mistral-community/pixtral-12b", },
     {"name": "seed-coder",       "tokt": TOKENIZER_TYPE.BPE, "repo": "https://huggingface.co/ByteDance-Seed/Seed-Coder-8B-Base", },
 ]
 
+# some models are known to be broken upstream, so we will skip them as exceptions
+pre_computed_hashes = [
+    # chatglm-bpe has 2 hashes, why?
+    {"name": "chatglm-bpe", "tokt": TOKENIZER_TYPE.BPE, "repo": "https://huggingface.co/THUDM/glm-4-9b-chat", "chkhsh": "b6e8e1518dc4305be2fe39c313ed643381c4da5db34a98f6a04c093f8afbe99b"},
+    {"name": "chatglm-bpe", "tokt": TOKENIZER_TYPE.BPE, "repo": "https://huggingface.co/THUDM/glm-4-9b-chat", "chkhsh": "81d72c7348a9f0ebe86f23298d37debe0a5e71149e29bd283904c02262b27516"},
+    {"name": "glm4", "tokt": TOKENIZER_TYPE.BPE, "repo": "https://huggingface.co/THUDM/glm-4-9b-hf", "chkhsh": "a1336059768a55c99a734006ffb02203cd450fed003e9a71886c88acf24fdbc2"},
+    {"name": "minerva-7b", "tokt": TOKENIZER_TYPE.BPE, "repo": "https://huggingface.co/sapienzanlp/Minerva-7B-base-v1.0", "chkhsh": "1431a23e583c97432bc230bff598d103ddb5a1f89960c8f1d1051aaa944d0b35"},
+]
+
 
 def download_file_with_auth(url, token, save_path):
     headers = {"Authorization": f"Bearer {token}"}
@@ -169,9 +189,29 @@ def download_model(model):
             if os.path.isfile(save_path):
                 logger.info(f"{name}: File {save_path} already exists - skipping")
                 continue
-            download_file_with_auth(f"{repo}/resolve/main/{file}", token, save_path)
+            download_file_with_auth(f"{repo}/resolve/main/{file}", hf_token, save_path)
 
 
+# get list of existing models and chkhsh from the convert_hf_to_gguf.py file
+# returns mapping res --> chkhsh
+def get_existing_models(convert_py):
+    pattern = r'if chkhsh == "([a-f0-9]{64})":\s*\n\s*.*\s*res = "([^"]+)"'
+    matches = re.findall(pattern, convert_py)
+    output = {}
+    for chkhsh, res in matches:
+        output[res] = chkhsh
+    return output
+
+
+existing_models = {}
+all_models = models.copy()
+if not args.full:
+    # Filter out models that already exist in convert_hf_to_gguf.py
+    existing_models = get_existing_models(convert_py)
+    all_models = models.copy()
+    models = [model for model in all_models if model["name"] not in existing_models]
+
+logging.info(f"Downloading {len(models)} models...")
 for model in models:
     try:
         download_model(model)
@@ -182,9 +222,10 @@ for model in models:
 # generate the source code for the convert_hf_to_gguf.py:get_vocab_base_pre() function:
 
 src_ifs = ""
-for model in models:
+for model in [*all_models, *pre_computed_hashes]:
     name = model["name"]
     tokt = model["tokt"]
+    chkhsh = model.get("chkhsh")
 
     if tokt == TOKENIZER_TYPE.SPM or tokt == TOKENIZER_TYPE.UGM:
         continue
@@ -195,7 +236,16 @@ for model in models:
         continue
 
     # create the tokenizer
+    if chkhsh is not None:
+        # if the model has a pre-computed hash, use it
+        logger.info(f"Using pre-computed hash for model {name}: {chkhsh}")
+    elif name in existing_models:
+        # if the model already exists in convert_hf_to_gguf.py, skip compute hash
+        chkhsh = existing_models[name]
+    else:
+        # otherwise, compute the hash of the tokenizer
         try:
+            logger.info(f"Loading tokenizer from {f'models/tokenizers/{name}'}...")
             if name == "t5":
                 tokenizer = AutoTokenizer.from_pretrained(f"models/tokenizers/{name}", use_fast=False)
             else:
@@ -271,8 +321,6 @@ src_func = f"""
         return res
 """
 
-convert_py_pth = pathlib.Path("convert_hf_to_gguf.py")
-convert_py = convert_py_pth.read_text(encoding="utf-8")
 convert_py = re.sub(
     r"(# Marker: Start get_vocab_base_pre)(.+?)( +# Marker: End get_vocab_base_pre)",
     lambda m: m.group(1) + src_func + m.group(3),
@@ -367,6 +415,10 @@ for model in models:
         logger.error(f"Failed to load tokenizer for model {name}. Error: {e}")
         continue  # Skip this model and continue with the next one in the loop
 
+    if not os.path.exists(f"models/ggml-vocab-{name}.gguf"):
+        logger.info(f"Skip vocab files for model {name}, no GGUF file found")
+        continue
+
     with open(f"models/ggml-vocab-{name}.gguf.inp", "w", encoding="utf-8") as f:
         for text in tests:
             f.write(f"{text}")

models/ggml-vocab-bert-bge.gguf.inp

@@ -1,6 +1,6 @@
 ied 4 ½ months
 __ggml_vocab_test__
-Führer
+Äpfel
 __ggml_vocab_test__
 
 __ggml_vocab_test__

models/ggml-vocab-bert-bge.gguf.out

@@ -1,5 +1,5 @@
  29464 2094 1018 1092 2706
- 11865 17875
+ 9706 7959 2140
 
 
 

models/ggml-vocab-chameleon.gguf.inp

@@ -1,112 +0,0 @@
-ied 4 ½ months
-__ggml_vocab_test__
-Führer
-__ggml_vocab_test__
-
-__ggml_vocab_test__
- 
-__ggml_vocab_test__
-  
-__ggml_vocab_test__
-   
-__ggml_vocab_test__
-	
-__ggml_vocab_test__
-
-
-__ggml_vocab_test__
-
-
-
-__ggml_vocab_test__
-
-
-
-
-__ggml_vocab_test__
-	
-
-__ggml_vocab_test__
-Hello world
-__ggml_vocab_test__
- Hello world
-__ggml_vocab_test__
-Hello World
-__ggml_vocab_test__
- Hello World
-__ggml_vocab_test__
- Hello World!
-__ggml_vocab_test__
-Hello, world!
-__ggml_vocab_test__
- Hello, world!
-__ggml_vocab_test__
- this is 🦙.cpp
-__ggml_vocab_test__
-w048 7tuijk dsdfhu
-__ggml_vocab_test__
-нещо на Български
-__ggml_vocab_test__
-កាន់តែពិសេសអាចខលចេញ
-__ggml_vocab_test__
-🚀 (normal) 😶‍🌫️ (multiple emojis concatenated) ✅ (only emoji that has its own token)
-__ggml_vocab_test__
-Hello
-__ggml_vocab_test__
- Hello
-__ggml_vocab_test__
-  Hello
-__ggml_vocab_test__
-   Hello
-__ggml_vocab_test__
-    Hello
-__ggml_vocab_test__
-    Hello
-    Hello
-__ggml_vocab_test__
- (
-__ggml_vocab_test__
-
- =
-__ggml_vocab_test__
-' era
-__ggml_vocab_test__
-Hello, y'all! How are you 😁 ?我想在apple工作1314151天～
-__ggml_vocab_test__
-!!!!!!
-__ggml_vocab_test__
-3
-__ggml_vocab_test__
-33
-__ggml_vocab_test__
-333
-__ggml_vocab_test__
-3333
-__ggml_vocab_test__
-33333
-__ggml_vocab_test__
-333333
-__ggml_vocab_test__
-3333333
-__ggml_vocab_test__
-33333333
-__ggml_vocab_test__
-333333333
-__ggml_vocab_test__
-Cửa Việt
-__ggml_vocab_test__
- discards
-__ggml_vocab_test__
-
- 
-
- 
-
-
- 	 		 	
-  
-   
-    
-     
-🚀 (normal) 😶‍🌫️ (multiple emojis concatenated) ✅ 🦙🦙 3 33 333 3333 33333 333333 3333333 33333333 3.3 3..3 3...3 កាន់តែពិសេសអាច😁 ?我想在apple工作1314151天～ ------======= нещо на Български ''''''```````""""......!!!!!!?????? I've been 'told he's there, 'RE you sure? 'M not sure I'll make it, 'D you like some tea? We'Ve a'lL
-__ggml_vocab_test__

models/ggml-vocab-chameleon.gguf.out

@@ -1,46 +0,0 @@
- 17245 16604 16403 16604 33583 18355
- 16421 51153
-
- 16604
- 16650
- 16650 16604
- 16581
- 16582
- 16582 16582
- 16582 16582 16582
- 16581 16582
- 31596 17394
- 34926 17394
- 31596 18671
- 34926 18671
- 34926 18671 16384
- 31596 16395 17394 16384
- 34926 16395 17394 16384
- 16811 16704 20410 16483 16631 16397 52854
- 16470 16399 16403 16407 16604 16406 35764 38185 51595 22592 26639
- 29479 23955 17012 20103 25527 27670 17408 19005 21473 24774
- 54254 42231 48084 29409 16617 61889 29409 16608 21954 16628 21954 16499 58445 29409 16607 58445 21954 16479 42231 21954 16611 21954 16607 21954 16633 21954 16611 29409 16607 21954 16615
- 52351 16604 16391 25825 16392 23686 16498 39161 18885 16618 16488 30853 16604 16391 54124 17153 25134 16656 18476 26169 16895 16392 62193 16611 16604 16391 24664 17153 57169 16721 16872 17073 17304 28729 16392
- 31596
- 34926
- 16650 31596
- 16650 34926
- 16696 31596
- 16696 31596 16582 16696 31596
- 16604 16391
- 16582 16604 16412
- 16390 22623
- 31596 16395 16712 16390 16828 16384 17674 16769 16732 23686 16607 16604 16414 24427 16623 41809 16495 28999 36469 45292 30197 16400 16402 16400 16403 16400 16404 16400 43969 65211 16636
- 16384 16384 16384 16384 16384 16384
- 16402
- 16402 16402
- 16402 16402 16402
- 16402 16402 16402 16402
- 16402 16402 16402 16402 16402
- 16402 16402 16402 16402 16402 16402
- 16402 16402 16402 16402 16402 16402 16402
- 16402 16402 16402 16402 16402 16402 16402 16402
- 16402 16402 16402 16402 16402 16402 16402 16402 16402
- 16418 19038 16639 16448 24315 33727 16467
- 18765 17981
- 16582 16604 16582 16582 16604 16582 16582 16582 16604 16581 16604 16581 16581 16604 16581 16582 16650 16582 16650 16604 16582 16696 16582 16696 16604 16582 52351 16604 16391 25825 16392 23686 16498 39161 18885 16618 16488 30853 16604 16391 54124 17153 25134 16656 18476 26169 16895 16392 62193 16611 20410 16483 16631 18885 16483 16631 16604 16402 16604 16402 16402 16604 16402 16402 16402 16604 16402 16402 16402 16402 16604 16402 16402 16402 16402 16402 16604 16402 16402 16402 16402 16402 16402 16604 16402 16402 16402 16402 16402 16402 16402 16604 16402 16402 16402 16402 16402 16402 16402 16402 16604 16402 16397 16402 16604 16402 16397 16397 16402 16604 16402 16397 16397 16397 16402 16604 54254 42231 48084 29409 16617 61889 29409 16608 21954 16628 21954 16499 58445 29409 16607 58445 21954 16479 42231 21954 16611 27683 16607 16604 16414 24427 16623 41809 16495 28999 36469 45292 30197 16400 16402 16400 16403 16400 16404 16400 43969 65211 16636 16604 16396 16396 16396 16396 16396 16396 16412 16412 16412 16412 16412 16412 16412 27268 23955 17012 20103 25527 27670 17408 19005 21473 24774 16604 16390 16390 16390 16390 16390 16390 16447 16447 16447 16447 16447 16447 16447 16385 16385 16385 16385 16397 16397 16397 16397 16397 16397 16384 16384 16384 16384 16384 16384 16414 16414 16414 16414 16414 16414 16687 16390 16690 16992 16604 16390 61797 16733 16390 16466 16986 16395 16604 16390 17879 16732 17811 16414 16604 16390 16428 16804 17811 16687 16390 16683 17190 16728 16395 16604 16390 16419 16732 16945 16991 25251 16414 17119 16390 38127 16641 16390 16459 16427

models/ggml-vocab-command-r.gguf.inp

@@ -1,6 +1,6 @@
 ied 4 ½ months
 __ggml_vocab_test__
-Führer
+Äpfel
 __ggml_vocab_test__
 
 __ggml_vocab_test__

models/ggml-vocab-command-r.gguf.out

@@ -1,5 +1,5 @@
  2536 228 27 228 22957 6983
- 45 193433
+ 90711 87 20910
 
  228
  1667

models/ggml-vocab-deepseek-coder.gguf.inp

@@ -1,6 +1,6 @@
 ied 4 ½ months
 __ggml_vocab_test__
-Führer
+Äpfel
 __ggml_vocab_test__
 
 __ggml_vocab_test__

models/ggml-vocab-deepseek-coder.gguf.out

@@ -1,5 +1,5 @@
  1050 207 19 207 19192 4217
- 37 32009 71 6247
+ 125 213 26862 282
 
  207
  243

models/ggml-vocab-deepseek-llm.gguf.inp

@@ -1,6 +1,6 @@
 ied 4 ½ months
 __ggml_vocab_test__
-Führer
+Äpfel
 __ggml_vocab_test__
 
 __ggml_vocab_test__

models/ggml-vocab-deepseek-llm.gguf.out

@@ -1,5 +1,5 @@
  1052 207 19 207 19109 4223
- 37 100014 71 6245
+ 82077 26723 282
 
  207
  243

models/ggml-vocab-deepseek-r1-qwen.gguf.inp

@@ -1,112 +0,0 @@
-ied 4 ½ months
-__ggml_vocab_test__
-Führer
-__ggml_vocab_test__
-
-__ggml_vocab_test__
- 
-__ggml_vocab_test__
-  
-__ggml_vocab_test__
-   
-__ggml_vocab_test__
-	
-__ggml_vocab_test__
-
-
-__ggml_vocab_test__
-
-
-
-__ggml_vocab_test__
-
-
-
-
-__ggml_vocab_test__
-	
-
-__ggml_vocab_test__
-Hello world
-__ggml_vocab_test__
- Hello world
-__ggml_vocab_test__
-Hello World
-__ggml_vocab_test__
- Hello World
-__ggml_vocab_test__
- Hello World!
-__ggml_vocab_test__
-Hello, world!
-__ggml_vocab_test__
- Hello, world!
-__ggml_vocab_test__
- this is 🦙.cpp
-__ggml_vocab_test__
-w048 7tuijk dsdfhu
-__ggml_vocab_test__
-нещо на Български
-__ggml_vocab_test__
-កាន់តែពិសេសអាចខលចេញ
-__ggml_vocab_test__
-🚀 (normal) 😶‍🌫️ (multiple emojis concatenated) ✅ (only emoji that has its own token)
-__ggml_vocab_test__
-Hello
-__ggml_vocab_test__
- Hello
-__ggml_vocab_test__
-  Hello
-__ggml_vocab_test__
-   Hello
-__ggml_vocab_test__
-    Hello
-__ggml_vocab_test__
-    Hello
-    Hello
-__ggml_vocab_test__
- (
-__ggml_vocab_test__
-
- =
-__ggml_vocab_test__
-' era
-__ggml_vocab_test__
-Hello, y'all! How are you 😁 ?我想在apple工作1314151天～
-__ggml_vocab_test__
-!!!!!!
-__ggml_vocab_test__
-3
-__ggml_vocab_test__
-33
-__ggml_vocab_test__
-333
-__ggml_vocab_test__
-3333
-__ggml_vocab_test__
-33333
-__ggml_vocab_test__
-333333
-__ggml_vocab_test__
-3333333
-__ggml_vocab_test__
-33333333
-__ggml_vocab_test__
-333333333
-__ggml_vocab_test__
-Cửa Việt
-__ggml_vocab_test__
- discards
-__ggml_vocab_test__
-
- 
-
- 
-
-
- 	 		 	
-  
-   
-    
-     
-🚀 (normal) 😶‍🌫️ (multiple emojis concatenated) ✅ 🦙🦙 3 33 333 3333 33333 333333 3333333 33333333 3.3 3..3 3...3 កាន់តែពិសេសអាច😁 ?我想在apple工作1314151天～ ------======= нещо на Български ''''''```````""""......!!!!!!?????? I've been 'told he's there, 'RE you sure? 'M not sure I'll make it, 'D you like some tea? We'Ve a'lL
-__ggml_vocab_test__

models/ggml-vocab-deepseek-r1-qwen.gguf.out

@@ -1,46 +0,0 @@
- 1122 220 19 220 26062 3951
- 37 50753 261
-
- 220
- 256
- 262
- 197
- 198
- 271
- 1406
- 1572
- 9707 1879
- 21927 1879
- 9707 4337
- 21927 4337
- 21927 4337 0
- 9707 11 1879 0
- 21927 11 1879 0
- 419 374 11162 99 247 13 10821
- 86 15 19 23 220 22 83 1963 41808 11472 2940 16739
- 78762 14144 1456 13073 63471 33594 3038 133178 79012
- 146394 97529 241 44258 233 146568 44258 224 147603 20879 115 146280 44258 223 146280 147272 97529 227 147805 148301 147270 44258 223 146848
- 145836 320 8252 8 26525 114 378 235 149921 30543 320 35673 99066 97534 8 25521 227 320 3243 42365 429 702 1181 1828 3950 8
- 9707
- 21927
- 220 21927
- 256 21927
- 262 21927
- 262 21927 198 262 21927
- 320
- 198 284
- 6 11385
- 9707 11 379 64848 0 2585 525 498 26525 223 937 104100 18493 22377 99257 16 18 16 19 16 20 16 35727 21216
- 17085 2928
- 18
- 18 18
- 18 18 18
- 18 18 18 18
- 18 18 18 18 18
- 18 18 18 18 18 18
- 18 18 18 18 18 18 18
- 18 18 18 18 18 18 18 18
- 18 18 18 18 18 18 18 18 18
- 34 90063 128324
- 2560 2347
- 198 4710 14731 65497 7847 1572 2303 78672 10947 145836 320 8252 8 26525 114 378 235 149921 30543 320 35673 99066 97534 8 25521 227 11162 99 247 149955 220 18 220 18 18 220 18 18 18 220 18 18 18 18 220 18 18 18 18 18 220 18 18 18 18 18 18 220 18 18 18 18 18 18 18 220 18 18 18 18 18 18 18 18 220 18 13 18 220 18 496 18 220 18 1112 18 220 146394 97529 241 44258 233 146568 44258 224 147603 20879 115 146280 44258 223 146280 147272 97529 227 144534 937 104100 18493 22377 99257 16 18 16 19 16 20 16 35727 21216 55460 53237 18658 14144 1456 13073 63471 33594 3038 133178 79012 3355 4605 4605 13874 13874 73594 3014 3014 28149 17085 2928 26610 7646 358 3003 1012 364 83 813 566 594 1052 11 364 787 498 2704 30 364 44 537 2704 358 3278 1281 432 11 364 35 498 1075 1045 15243 30 1205 6 42612 264 63866 43

models/ggml-vocab-falcon.gguf.inp

@@ -1,6 +1,6 @@
 ied 4 ½ months
 __ggml_vocab_test__
-Führer
+Äpfel
 __ggml_vocab_test__
 
 __ggml_vocab_test__

models/ggml-vocab-falcon.gguf.out

@@ -1,5 +1,5 @@
  878 204 31 3068 133 2137
- 28611 132 30042
+ 34502 18614 286
 
  204
  258

models/ggml-vocab-gpt-2.gguf.inp

@@ -1,6 +1,6 @@
 ied 4 ½ months
 __ggml_vocab_test__
-Führer
+Äpfel
 __ggml_vocab_test__
 
 __ggml_vocab_test__

models/ggml-vocab-gpt-2.gguf.out

@@ -1,5 +1,5 @@
  798 604 25208 1933
- 37 9116 71 11751
+ 127 226 79 69 417
 
  220
  220 220

models/ggml-vocab-gpt-4o.gguf.inp

@@ -1,112 +0,0 @@
-ied 4 ½ months
-__ggml_vocab_test__
-Führer
-__ggml_vocab_test__
-
-__ggml_vocab_test__
- 
-__ggml_vocab_test__
-  
-__ggml_vocab_test__
-   
-__ggml_vocab_test__
-	
-__ggml_vocab_test__
-
-
-__ggml_vocab_test__
-
-
-
-__ggml_vocab_test__
-
-
-
-
-__ggml_vocab_test__
-	
-
-__ggml_vocab_test__
-Hello world
-__ggml_vocab_test__
- Hello world
-__ggml_vocab_test__
-Hello World
-__ggml_vocab_test__
- Hello World
-__ggml_vocab_test__
- Hello World!
-__ggml_vocab_test__
-Hello, world!
-__ggml_vocab_test__
- Hello, world!
-__ggml_vocab_test__
- this is 🦙.cpp
-__ggml_vocab_test__
-w048 7tuijk dsdfhu
-__ggml_vocab_test__
-нещо на Български
-__ggml_vocab_test__
-កាន់តែពិសេសអាចខលចេញ
-__ggml_vocab_test__
-🚀 (normal) 😶‍🌫️ (multiple emojis concatenated) ✅ (only emoji that has its own token)
-__ggml_vocab_test__
-Hello
-__ggml_vocab_test__
- Hello
-__ggml_vocab_test__
-  Hello
-__ggml_vocab_test__
-   Hello
-__ggml_vocab_test__
-    Hello
-__ggml_vocab_test__
-    Hello
-    Hello
-__ggml_vocab_test__
- (
-__ggml_vocab_test__
-
- =
-__ggml_vocab_test__
-' era
-__ggml_vocab_test__
-Hello, y'all! How are you 😁 ?我想在apple工作1314151天～
-__ggml_vocab_test__
-!!!!!!
-__ggml_vocab_test__
-3
-__ggml_vocab_test__
-33
-__ggml_vocab_test__
-333
-__ggml_vocab_test__
-3333
-__ggml_vocab_test__
-33333
-__ggml_vocab_test__
-333333
-__ggml_vocab_test__
-3333333
-__ggml_vocab_test__
-33333333
-__ggml_vocab_test__
-333333333
-__ggml_vocab_test__
-Cửa Việt
-__ggml_vocab_test__
- discards
-__ggml_vocab_test__
-
- 
-
- 
-
-
- 	 		 	
-  
-   
-    
-     
-🚀 (normal) 😶‍🌫️ (multiple emojis concatenated) ✅ 🦙🦙 3 33 333 3333 33333 333333 3333333 33333333 3.3 3..3 3...3 កាន់តែពិសេសអាច😁 ?我想在apple工作1314151天～ ------======= нещо на Български ''''''```````""""......!!!!!!?????? I've been 'told he's there, 'RE you sure? 'M not sure I'll make it, 'D you like some tea? We'Ve a'lL
-__ggml_vocab_test__

models/ggml-vocab-gpt-4o.gguf.out

@@ -1,46 +0,0 @@
- 1165 220 19 220 27124 5503
- 37 19194 259
-
- 220
- 256
- 271
- 197
- 198
- 279
- 2499
- 2775
- 13225 2375
- 32949 2375
- 13225 5922
- 32949 5922
- 32949 5922 0
- 13225 11 2375 0
- 32949 11 2375 0
- 495 382 9552 99 247 13 17159
- 86 45404 220 22 10191 2852 22924 4750 6916
- 3907 53641 1235 185386 8118
- 11400 107516 15867 20804 22851 134178 77431 32010 104312 37984 16329 27751 89335
- 112927 222 350 14559 8 22861 114 2524 64364 104 15148 350 76466 166700 121942 780 8 91349 350 7393 74471 484 853 1617 2316 6602 8
- 13225
- 32949
- 220 32949
- 256 32949
- 271 32949
- 271 32949 198 271 32949
- 350
- 198 314
- 6 6837
- 13225 11 342 70653 0 3253 553 481 22861 223 1423 7522 18165 2178 34058 22369 16412 32999 16 867 8208
- 147475
- 18
- 2546
- 15517
- 15517 18
- 15517 2546
- 15517 15517
- 15517 15517 18
- 15517 15517 2546
- 15517 15517 15517
- 34 60213 53904
- 2960 3098
- 126470 25980 160432 16609 2775 4066 172261 19432 112927 222 350 14559 8 22861 114 2524 64364 104 15148 350 76466 166700 121942 780 8 91349 9552 99 247 4103 99 247 220 18 220 2546 220 15517 220 15517 18 220 15517 2546 220 15517 15517 220 15517 15517 18 220 15517 15517 2546 220 18 13 18 220 18 485 18 220 18 1008 18 44735 107516 15867 20804 22851 134178 77431 32010 104312 156437 1423 7522 18165 2178 34058 22369 16412 32999 16 867 8208 105024 106657 1967 53641 1235 185386 8118 22434 39336 26178 26178 168394 194663 27271 147475 25883 6961 9790 1339 461 83 1280 19016 1354 11 461 1099 481 3239 30 461 44 625 3239 17291 1520 480 11 461 35 481 1299 1236 17966 30 1416 6 27493 261 54602 43

models/ggml-vocab-llama-bpe.gguf.inp

@@ -1,6 +1,6 @@
 ied 4 ½ months
 __ggml_vocab_test__
-Führer
+Äpfel
 __ggml_vocab_test__
 
 __ggml_vocab_test__

models/ggml-vocab-llama-bpe.gguf.out

@@ -1,5 +1,5 @@
  1142 220 19 220 27154 4038
- 37 51853 261
+ 88075 16276 301
 
  220
  256

models/ggml-vocab-llama-spm.gguf.inp

@@ -1,6 +1,6 @@
 ied 4 ½ months
 __ggml_vocab_test__
-Führer
+Äpfel
 __ggml_vocab_test__
 
 __ggml_vocab_test__

models/ggml-vocab-llama-spm.gguf.out

@@ -1,5 +1,5 @@
  474 287 29871 29946 29871 30226 7378
- 383 4000 261
+ 11585 7810 295
 
  259
  1678

models/ggml-vocab-llama4.gguf.inp

@@ -1,112 +0,0 @@
-ied 4 ½ months
-__ggml_vocab_test__
-Führer
-__ggml_vocab_test__
-
-__ggml_vocab_test__
- 
-__ggml_vocab_test__
-  
-__ggml_vocab_test__
-   
-__ggml_vocab_test__
-	
-__ggml_vocab_test__
-
-
-__ggml_vocab_test__
-
-
-
-__ggml_vocab_test__
-
-
-
-
-__ggml_vocab_test__
-	
-
-__ggml_vocab_test__
-Hello world
-__ggml_vocab_test__
- Hello world
-__ggml_vocab_test__
-Hello World
-__ggml_vocab_test__
- Hello World
-__ggml_vocab_test__
- Hello World!
-__ggml_vocab_test__
-Hello, world!
-__ggml_vocab_test__
- Hello, world!
-__ggml_vocab_test__
- this is 🦙.cpp
-__ggml_vocab_test__
-w048 7tuijk dsdfhu
-__ggml_vocab_test__
-нещо на Български
-__ggml_vocab_test__
-កាន់តែពិសេសអាចខលចេញ
-__ggml_vocab_test__
-🚀 (normal) 😶‍🌫️ (multiple emojis concatenated) ✅ (only emoji that has its own token)
-__ggml_vocab_test__
-Hello
-__ggml_vocab_test__
- Hello
-__ggml_vocab_test__
-  Hello
-__ggml_vocab_test__
-   Hello
-__ggml_vocab_test__
-    Hello
-__ggml_vocab_test__
-    Hello
-    Hello
-__ggml_vocab_test__
- (
-__ggml_vocab_test__
-
- =
-__ggml_vocab_test__
-' era
-__ggml_vocab_test__
-Hello, y'all! How are you 😁 ?我想在apple工作1314151天～
-__ggml_vocab_test__
-!!!!!!
-__ggml_vocab_test__
-3
-__ggml_vocab_test__
-33
-__ggml_vocab_test__
-333
-__ggml_vocab_test__
-3333
-__ggml_vocab_test__
-33333
-__ggml_vocab_test__
-333333
-__ggml_vocab_test__
-3333333
-__ggml_vocab_test__
-33333333
-__ggml_vocab_test__
-333333333
-__ggml_vocab_test__
-Cửa Việt
-__ggml_vocab_test__
- discards
-__ggml_vocab_test__
-
- 
-
- 
-
-
- 	 		 	
-  
-   
-    
-     
-🚀 (normal) 😶‍🌫️ (multiple emojis concatenated) ✅ 🦙🦙 3 33 333 3333 33333 333333 3333333 33333333 3.3 3..3 3...3 កាន់តែពិសេសអាច😁 ?我想在apple工作1314151天～ ------======= нещо на Български ''''''```````""""......!!!!!!?????? I've been 'told he's there, 'RE you sure? 'M not sure I'll make it, 'D you like some tea? We'Ve a'lL
-__ggml_vocab_test__

models/ggml-vocab-llama4.gguf.out

@@ -1,46 +0,0 @@
- 1190 220 32 220 18215 7112
- 50 16800 258
-
- 220
- 256
- 277
- 197
- 198
- 368
- 2946
- 3271
- 19873 3817
- 39715 3817
- 19873 7353
- 39715 7353
- 39715 7353 13
- 19873 24 3817 13
- 39715 24 3817 13
- 544 373 9522 112 247 26 36315
- 99 39923 220 35 9607 21498 21470 3679 9433
- 1595 7653 633 79829 34051 1636
- 8755 102595 115960 21125 148305 96819 102816 39048 14105 22528 160234
- 114590 222 330 14879 21 51358 127 12817 93293 117 24204 330 68239 881 120327 170428 21 89101 330 7384 88230 511 947 1492 3742 7233 21
- 19873
- 39715
- 220 39715
- 256 39715
- 277 39715
- 277 39715 198 277 39715
- 330
- 198 319
- 19 7359
- 19873 24 386 87799 13 2403 583 650 51358 223 1663 155736 1522 42056 7544 13336 28785 29 4412 20645
- 17931 4959
- 31
- 1922
- 12325
- 12325 31
- 12325 1922
- 12325 12325
- 12325 12325 31
- 12325 12325 1922
- 12325 12325 12325
- 47 19811 12077
- 3260 3579
- 198 7283 51499 191231 20192 3271 3322 9287 2143 17860 114590 222 330 14879 21 51358 127 12817 93293 117 24204 330 68239 881 120327 170428 21 89101 9522 112 247 172394 247 220 31 220 1922 220 12325 220 12325 31 220 12325 1922 220 12325 12325 220 12325 12325 31 220 12325 12325 1922 220 31 26 31 220 31 396 31 220 31 1043 31 117131 102595 115960 21125 148305 96819 102816 80883 223 1663 155736 1522 42056 7544 13336 28785 29 4412 20645 79745 150278 117079 633 79829 34051 1636 25611 41990 109428 1488 91054 24072 17931 4959 29795 9296 16517 1806 481 96 1386 36633 1609 24 481 1109 650 5074 43 481 57 702 5074 27088 2170 536 24 481 48 650 1933 1696 30262 43 1665 19 32818 262 27236 56

models/ggml-vocab-mpt.gguf.inp

@@ -1,6 +1,6 @@
 ied 4 ½ months
 __ggml_vocab_test__
-Führer
+Äpfel
 __ggml_vocab_test__
 
 __ggml_vocab_test__

models/ggml-vocab-mpt.gguf.out

@@ -1,5 +1,5 @@
  728 577 24142 2607
- 39 26288 6554
+ 37515 18569 293
 
  209
  50276

models/ggml-vocab-nomic-bert-moe.gguf.inp

@@ -1,112 +0,0 @@
-ied 4 ½ months
-__ggml_vocab_test__
-Führer
-__ggml_vocab_test__
-
-__ggml_vocab_test__
- 
-__ggml_vocab_test__
-  
-__ggml_vocab_test__
-   
-__ggml_vocab_test__
-	
-__ggml_vocab_test__
-
-
-__ggml_vocab_test__
-
-
-
-__ggml_vocab_test__
-
-
-
-
-__ggml_vocab_test__
-	
-
-__ggml_vocab_test__
-Hello world
-__ggml_vocab_test__
- Hello world
-__ggml_vocab_test__
-Hello World
-__ggml_vocab_test__
- Hello World
-__ggml_vocab_test__
- Hello World!
-__ggml_vocab_test__
-Hello, world!
-__ggml_vocab_test__
- Hello, world!
-__ggml_vocab_test__
- this is 🦙.cpp
-__ggml_vocab_test__
-w048 7tuijk dsdfhu
-__ggml_vocab_test__
-нещо на Български
-__ggml_vocab_test__
-កាន់តែពិសេសអាចខលចេញ
-__ggml_vocab_test__
-🚀 (normal) 😶‍🌫️ (multiple emojis concatenated) ✅ (only emoji that has its own token)
-__ggml_vocab_test__
-Hello
-__ggml_vocab_test__
- Hello
-__ggml_vocab_test__
-  Hello
-__ggml_vocab_test__
-   Hello
-__ggml_vocab_test__
-    Hello
-__ggml_vocab_test__
-    Hello
-    Hello
-__ggml_vocab_test__
- (
-__ggml_vocab_test__
-
- =
-__ggml_vocab_test__
-' era
-__ggml_vocab_test__
-Hello, y'all! How are you 😁 ?我想在apple工作1314151天～
-__ggml_vocab_test__
-!!!!!!
-__ggml_vocab_test__
-3
-__ggml_vocab_test__
-33
-__ggml_vocab_test__
-333
-__ggml_vocab_test__
-3333
-__ggml_vocab_test__
-33333
-__ggml_vocab_test__
-333333
-__ggml_vocab_test__
-3333333
-__ggml_vocab_test__
-33333333
-__ggml_vocab_test__
-333333333
-__ggml_vocab_test__
-Cửa Việt
-__ggml_vocab_test__
- discards
-__ggml_vocab_test__
-
- 
-
- 
-
-
- 	 		 	
-  
-   
-    
-     
-🚀 (normal) 😶‍🌫️ (multiple emojis concatenated) ✅ 🦙🦙 3 33 333 3333 33333 333333 3333333 33333333 3.3 3..3 3...3 កាន់តែពិសេសអាច😁 ?我想在apple工作1314151天～ ------======= нещо на Български ''''''```````""""......!!!!!!?????? I've been 'told he's there, 'RE you sure? 'M not sure I'll make it, 'D you like some tea? We'Ve a'lL
-__ggml_vocab_test__

models/ggml-vocab-nomic-bert-moe.gguf.out

@@ -1,46 +0,0 @@
- 17 297 201 78660 21775
- 72805 4097 56
-
-
-
-
-
-
-
-
-
- 35378 8999
- 35378 8999
- 35378 6661
- 35378 6661
- 35378 6661 38
- 35378 4 8999 38
- 35378 4 8999 38
- 903 83 6 3 5 238 6366
- 148 7709 1019 361 458 134362 104 7 71 420 1132
- 14271 29 117152
- 6 149561 78270 48967 64254 7616 81705
- 6 247206 15 33176 16 6 247442 6 3 15755 15 144227 8705 18255 40292 158 4460 33 27686 16 6 142325 15 191 538 28 121505 450 1556 6863 10002 47 1098 16
- 35378
- 35378
- 35378
- 35378
- 35378
- 35378 35378
- 15
- 2203
- 242 1615
- 35378 4 113 25 5584 38 11249 621 398 6 201344 705 23638 213 9007 133 1879 2681 2592 135224 1906 6087
- 6 90827
- 138
- 3912
- 6 66000
- 138 66000
- 3912 66000
- 6 66000 66000
- 138 66000 66000
- 3912 66000 66000
- 6 66000 66000 66000
- 199152 3763
- 17116 99397
- 6 247206 15 33176 16 6 247442 6 3 15755 15 144227 8705 18255 40292 158 4460 33 27686 16 6 142325 6 3 138 3912 6 66000 138 66000 3912 66000 6 66000 66000 138 66000 66000 3912 66000 66000 80308 1031 5 363 138 27 363 6 149561 78270 48967 201344 705 23638 213 9007 133 1879 2681 2592 135224 1906 6087 6 110405 1369 69112 69112 69112 14271 29 117152 5106 4765 4765 1135 164721 164721 164721 58 58 58 58 2551 90827 32 85908 87 25 272 2809 242 18 18345 764 25 7 2685 4 242 11766 398 9077 32 242 594 959 9077 87 25 1181 3249 442 4 242 397 398 1884 3060 26156 32 1401 25 26455 10 25 141 866

models/ggml-vocab-phi-3.gguf.inp

@@ -1,6 +1,6 @@
 ied 4 ½ months
 __ggml_vocab_test__
-Führer
+Äpfel
 __ggml_vocab_test__
 
 __ggml_vocab_test__

models/ggml-vocab-phi-3.gguf.out

@@ -1,5 +1,5 @@
  474 287 29871 29946 29871 30226 7378
- 383 4000 261
+ 11585 7810 295
 
  259
  1678

models/ggml-vocab-pixtral.gguf.inp

@@ -1,112 +0,0 @@
-ied 4 ½ months
-__ggml_vocab_test__
-Führer
-__ggml_vocab_test__
-
-__ggml_vocab_test__
- 
-__ggml_vocab_test__
-  
-__ggml_vocab_test__
-   
-__ggml_vocab_test__
-	
-__ggml_vocab_test__
-
-
-__ggml_vocab_test__
-
-
-
-__ggml_vocab_test__
-
-
-
-
-__ggml_vocab_test__
-	
-
-__ggml_vocab_test__
-Hello world
-__ggml_vocab_test__
- Hello world
-__ggml_vocab_test__
-Hello World
-__ggml_vocab_test__
- Hello World
-__ggml_vocab_test__
- Hello World!
-__ggml_vocab_test__
-Hello, world!
-__ggml_vocab_test__
- Hello, world!
-__ggml_vocab_test__
- this is 🦙.cpp
-__ggml_vocab_test__
-w048 7tuijk dsdfhu
-__ggml_vocab_test__
-нещо на Български
-__ggml_vocab_test__
-កាន់តែពិសេសអាចខលចេញ
-__ggml_vocab_test__
-🚀 (normal) 😶‍🌫️ (multiple emojis concatenated) ✅ (only emoji that has its own token)
-__ggml_vocab_test__
-Hello
-__ggml_vocab_test__
- Hello
-__ggml_vocab_test__
-  Hello
-__ggml_vocab_test__
-   Hello
-__ggml_vocab_test__
-    Hello
-__ggml_vocab_test__
-    Hello
-    Hello
-__ggml_vocab_test__
- (
-__ggml_vocab_test__
-
- =
-__ggml_vocab_test__
-' era
-__ggml_vocab_test__
-Hello, y'all! How are you 😁 ?我想在apple工作1314151天～
-__ggml_vocab_test__
-!!!!!!
-__ggml_vocab_test__
-3
-__ggml_vocab_test__
-33
-__ggml_vocab_test__
-333
-__ggml_vocab_test__
-3333
-__ggml_vocab_test__
-33333
-__ggml_vocab_test__
-333333
-__ggml_vocab_test__
-3333333
-__ggml_vocab_test__
-33333333
-__ggml_vocab_test__
-333333333
-__ggml_vocab_test__
-Cửa Việt
-__ggml_vocab_test__
- discards
-__ggml_vocab_test__
-
- 
-
- 
-
-
- 	 		 	
-  
-   
-    
-     
-🚀 (normal) 😶‍🌫️ (multiple emojis concatenated) ✅ 🦙🦙 3 33 333 3333 33333 333333 3333333 33333333 3.3 3..3 3...3 កាន់តែពិសេសអាច😁 ?我想在apple工作1314151天～ ------======= нещо на Български ''''''```````""""......!!!!!!?????? I've been 'told he's there, 'RE you sure? 'M not sure I'll make it, 'D you like some tea? We'Ve a'lL
-__ggml_vocab_test__

models/ggml-vocab-pixtral.gguf.out

@@ -1,46 +0,0 @@
- 2014 1032 1052 1032 28504 6972
- 1070 7088 1258
-
- 1032
- 1256
- 1293
- 1009
- 1010
- 1267
- 4688
- 1009 1010
- 22177 4304
- 45383 4304
- 22177 5325
- 45383 5325
- 45383 5325 1033
- 22177 1044 4304 1033
- 45383 1044 4304 1033
- 1593 1395 119685 1166 1153 1046 51228
- 1119 1048 1052 1056 1032 1055 17391 23216 30203 7785 17279
- 3337 30757 1902 4200 63073 3671
- 1225 1158 1128 1225 1158 1182 1225 1158 1147 1225 1159 1139 1225 1158 1143 1225 1159 1130 1225 1158 1150 1225 1158 1183 1225 1158 1159 1225 21359 1225 1158 1159 1225 1158 1162 1225 1158 1182 1225 1158 1133 1225 1158 1129 1225 1158 1155 1225 1158 1133 1225 21359 1225 1158 1137
- 1240 1159 1154 1128 1319 13052 1041 119685 1152 1182 29568 1240 1159 1140 1171 1239 1184 1143 1319 88181 1873 3659 1275 56421 1621 1041 126241 1133 1319 11234 1873 26303 1455 1934 2246 3754 10835 1041
- 22177
- 45383
- 1032 45383
- 1256 45383
- 1293 45383
- 1293 45383 1010 1293 45383
- 1319
- 1010 1376
- 1039 4033
- 22177 1044 1404 48054 1033 3075 1584 1636 119685 1152 1129 3082 26060 2998 63614 82278 1049 1051 1049 1052 1049 1053 1049 6434 6749
- 7290 7290 7290
- 1051
- 1051 1051
- 1051 1051 1051
- 1051 1051 1051 1051
- 1051 1051 1051 1051 1051
- 1051 1051 1051 1051 1051 1051
- 1051 1051 1051 1051 1051 1051 1051
- 1051 1051 1051 1051 1051 1051 1051 1051
- 1051 1051 1051 1051 1051 1051 1051 1051 1051
- 1067 59503 28783
- 3724 4058
- 1010 1032 1267 1032 4688 1032 17152 1458 29356 1010 1256 1010 1293 1010 1260 1010 1652 1010 1240 1159 1154 1128 1319 13052 1041 119685 1152 1182 29568 1240 1159 1140 1171 1239 1184 1143 1319 88181 1873 3659 1275 56421 1621 1041 126241 1133 119685 1166 1153 1240 1159 1166 1153 1032 1051 1032 1051 1051 1032 1051 1051 1051 1032 1051 1051 1051 1051 1032 1051 1051 1051 1051 1051 1032 1051 1051 1051 1051 1051 1051 1032 1051 1051 1051 1051 1051 1051 1051 1032 1051 1051 1051 1051 1051 1051 1051 1051 1032 1051 1046 1051 1032 1051 1791 1051 1032 1051 2880 1051 71881 1158 1128 1225 1158 1182 1225 1158 1147 1225 1159 1139 1225 1158 1143 1225 1159 1130 1225 1158 1150 1225 1158 1183 1225 1158 1159 1225 21359 1225 1158 1159 1225 1158 1162 1225 1158 1182 1225 1158 1133 1240 1159 1152 1129 3082 26060 2998 63614 82278 1049 1051 1049 1052 1049 1053 1049 6434 6749 45577 1045 6626 43555 2843 30757 1902 4200 63073 3671 14931 20040 20040 1657 1657 1975 14135 14135 83923 7290 7290 7290 45509 45509 45509 1362 6483 2151 1576 1116 2189 1514 1681 2156 1044 1576 3609 1636 5257 1063 1576 1077 1605 5257 1362 7534 3180 1494 1044 1576 1068 1636 2479 2269 26883 1063 2837 1039 45654 1261 54297 1076

models/ggml-vocab-qwen2.gguf.inp

@@ -1,6 +1,6 @@
 ied 4 ½ months
 __ggml_vocab_test__
-Führer
+Äpfel
 __ggml_vocab_test__
 
 __ggml_vocab_test__

models/ggml-vocab-qwen2.gguf.out

@@ -1,5 +1,5 @@
  1122 220 19 220 26062 3951
- 37 50753 261
+ 86975 15897 301
 
  220
  256

models/ggml-vocab-refact.gguf.inp

@@ -1,6 +1,6 @@
 ied 4 ½ months
 __ggml_vocab_test__
-Führer
+Äpfel
 __ggml_vocab_test__
 
 __ggml_vocab_test__

models/ggml-vocab-refact.gguf.out

@@ -1,5 +1,5 @@
  4833 225 38 225 143 140 17723
- 56 2006 3935 265
+ 144 231 7132 342
 
  225
  261

models/ggml-vocab-roberta-bpe.gguf.inp

@@ -1,112 +0,0 @@
-ied 4 ½ months
-__ggml_vocab_test__
-Führer
-__ggml_vocab_test__
-
-__ggml_vocab_test__
- 
-__ggml_vocab_test__
-  
-__ggml_vocab_test__
-   
-__ggml_vocab_test__
-	
-__ggml_vocab_test__
-
-
-__ggml_vocab_test__
-
-
-
-__ggml_vocab_test__
-
-
-
-
-__ggml_vocab_test__
-	
-
-__ggml_vocab_test__
-Hello world
-__ggml_vocab_test__
- Hello world
-__ggml_vocab_test__
-Hello World
-__ggml_vocab_test__
- Hello World
-__ggml_vocab_test__
- Hello World!
-__ggml_vocab_test__
-Hello, world!
-__ggml_vocab_test__
- Hello, world!
-__ggml_vocab_test__
- this is 🦙.cpp
-__ggml_vocab_test__
-w048 7tuijk dsdfhu
-__ggml_vocab_test__
-нещо на Български
-__ggml_vocab_test__
-កាន់តែពិសេសអាចខលចេញ
-__ggml_vocab_test__
-🚀 (normal) 😶‍🌫️ (multiple emojis concatenated) ✅ (only emoji that has its own token)
-__ggml_vocab_test__
-Hello
-__ggml_vocab_test__
- Hello
-__ggml_vocab_test__
-  Hello
-__ggml_vocab_test__
-   Hello
-__ggml_vocab_test__
-    Hello
-__ggml_vocab_test__
-    Hello
-    Hello
-__ggml_vocab_test__
- (
-__ggml_vocab_test__
-
- =
-__ggml_vocab_test__
-' era
-__ggml_vocab_test__
-Hello, y'all! How are you 😁 ?我想在apple工作1314151天～
-__ggml_vocab_test__
-!!!!!!
-__ggml_vocab_test__
-3
-__ggml_vocab_test__
-33
-__ggml_vocab_test__
-333
-__ggml_vocab_test__
-3333
-__ggml_vocab_test__
-33333
-__ggml_vocab_test__
-333333
-__ggml_vocab_test__
-3333333
-__ggml_vocab_test__
-33333333
-__ggml_vocab_test__
-333333333
-__ggml_vocab_test__
-Cửa Việt
-__ggml_vocab_test__
- discards
-__ggml_vocab_test__
-
- 
-
- 
-
-
- 	 		 	
-  
-   
-    
-     
-🚀 (normal) 😶‍🌫️ (multiple emojis concatenated) ✅ 🦙🦙 3 33 333 3333 33333 333333 3333333 33333333 3.3 3..3 3...3 កាន់តែពិសេសអាច😁 ?我想在apple工作1314151天～ ------======= нещо на Български ''''''```````""""......!!!!!!?????? I've been 'told he's there, 'RE you sure? 'M not sure I'll make it, 'D you like some tea? We'Ve a'lL
-__ggml_vocab_test__

models/ggml-vocab-roberta-bpe.gguf.out

@@ -1,46 +0,0 @@
- 2550 204 18430 377
- 597 2768 298 8564
-
- 1437
- 1437 1437
- 1437 1437 1437
- 50117
- 50118
- 50140
- 50140 50118
- 50117 50118
- 31414 232
- 20920 232
- 31414 623
- 20920 623
- 20920 623 328
- 31414 6 232 328
- 20920 6 232 328
- 42 16 8103 18164 27 4 49317
- 605 40976 262 10109 18474 385 29 36807 6455
- 36765 25482 22063 23171 34251 18697 10809 26161 18697 3602 22063 27969 40966 25417 15264 26161 24269 36709 41171 35328
- 1376 17772 7471 1376 17772 19002 1376 17772 9085 1376 4333 13859 1376 17772 9357 1376 4333 9264 1376 17772 25448 1376 17772 18400 1376 17772 4333 1376 4333 10172 1376 17772 4333 1376 17772 7258 1376 17772 19002 1376 17772 5782 1376 17772 10172 1376 17772 3726 1376 17772 5782 1376 4333 10172 1376 17772 23171
- 6569 15113 7471 36 21113 43 17841 19002 17 8384 6569 14285 4958 12605 36 34654 2841 4203 354 10146 26511 1070 43 36174 5782 36 8338 21554 14 34 63 308 19233 43
- 31414
- 20920
- 1437 20920
- 1437 1437 20920
- 1437 1437 1437 20920
- 1437 1437 1437 20920 50118 1437 1437 1437 20920
- 36
- 50118 5457
- 108 3567
- 31414 6 1423 108 1250 328 1336 32 47 17841 10172 17487 47876 3602 48617 15264 46537 11423 27326 48494 8210 49233 1558 1570 27761 49429 43251 10809 17772
- 32376 12846
- 246
- 3103
- 25631
- 46152
- 3103 25631
- 46152 3103
- 46152 25631
- 46152 46152
- 46152 3103 25631
- 347 1376 2023 12410 102 16376 1376 2023 6382 90
- 9553 5954
- 50118 1437 50140 1437 50140 50118 1437 50117 1437 50117 50117 1437 50117 50118 1437 1437 50118 1437 1437 1437 50118 1437 1437 1437 1437 50118 1437 1437 1437 1437 1437 50118 6569 15113 7471 36 21113 43 17841 19002 17 8384 6569 14285 4958 12605 36 34654 2841 4203 354 10146 26511 1070 43 36174 5782 8103 18164 27 6569 18164 27 155 2357 30242 155 25631 30242 3103 30242 25631 30242 46152 30242 3103 25631 155 4 246 155 7586 246 155 734 246 25974 17772 7471 1376 17772 19002 1376 17772 9085 1376 4333 13859 1376 17772 9357 1376 4333 9264 1376 17772 25448 1376 17772 18400 1376 17772 4333 1376 4333 10172 1376 17772 4333 1376 17772 7258 1376 17772 19002 1376 17772 5782 18636 10172 17487 47876 3602 48617 15264 46537 11423 27326 48494 8210 49233 1558 1570 27761 49429 43251 10809 17772 36738 48332 47463 18697 10809 25482 22063 23171 34251 18697 10809 26161 18697 3602 22063 27969 40966 25417 15264 26161 24269 36709 41171 35328 128 49690 108 49972 49519 12905 48149 48149 43796 32376 12846 27282 28749 38 348 57 128 41042 37 18 89 6 128 4629 47 686 116 128 448 45 686 38 581 146 24 6 128 495 47 101 103 6845 116 166 108 30660 10 108 462 574

models/ggml-vocab-starcoder.gguf.inp

@@ -1,6 +1,6 @@
 ied 4 ½ months
 __ggml_vocab_test__
-Führer
+Äpfel
 __ggml_vocab_test__
 
 __ggml_vocab_test__

models/ggml-vocab-starcoder.gguf.out

@@ -1,5 +1,5 @@
  4850 244 57 244 162 159 17722
- 75 2022 3943 284
+ 163 250 7146 361
 
  244
  280

tests/CMakeLists.txt

@@ -92,12 +92,14 @@ llama_test(test-tokenizer-0 NAME test-tokenizer-0-gpt-2             ARGS ${CMAKE
 llama_test(test-tokenizer-0 NAME test-tokenizer-0-llama-bpe         ARGS ${CMAKE_CURRENT_SOURCE_DIR}/../models/ggml-vocab-llama-bpe.gguf)
 llama_test(test-tokenizer-0 NAME test-tokenizer-0-llama-spm         ARGS ${CMAKE_CURRENT_SOURCE_DIR}/../models/ggml-vocab-llama-spm.gguf)
 llama_test(test-tokenizer-0 NAME test-tokenizer-0-mpt               ARGS ${CMAKE_CURRENT_SOURCE_DIR}/../models/ggml-vocab-mpt.gguf)
-llama_test(test-tokenizer-0 NAME test-tokenizer-0-nomic-bert-moe    ARGS ${CMAKE_CURRENT_SOURCE_DIR}/../models/ggml-vocab-nomic-bert-moe.gguf)
 llama_test(test-tokenizer-0 NAME test-tokenizer-0-phi-3             ARGS ${CMAKE_CURRENT_SOURCE_DIR}/../models/ggml-vocab-phi-3.gguf)
 llama_test(test-tokenizer-0 NAME test-tokenizer-0-qwen2             ARGS ${CMAKE_CURRENT_SOURCE_DIR}/../models/ggml-vocab-qwen2.gguf)
 llama_test(test-tokenizer-0 NAME test-tokenizer-0-refact            ARGS ${CMAKE_CURRENT_SOURCE_DIR}/../models/ggml-vocab-refact.gguf)
 llama_test(test-tokenizer-0 NAME test-tokenizer-0-starcoder         ARGS ${CMAKE_CURRENT_SOURCE_DIR}/../models/ggml-vocab-starcoder.gguf)
 
+# TODO: missing HF tokenizer for this model in convert_hf_to_gguf_update.py, see https://github.com/ggml-org/llama.cpp/pull/13847
+# llama_test(test-tokenizer-0 NAME test-tokenizer-0-nomic-bert-moe    ARGS ${CMAKE_CURRENT_SOURCE_DIR}/../models/ggml-vocab-nomic-bert-moe.gguf)
+
 if (LLAMA_LLGUIDANCE)
     llama_build_and_test(test-grammar-llguidance.cpp ARGS ${CMAKE_CURRENT_SOURCE_DIR}/../models/ggml-vocab-llama-bpe.gguf)
 endif ()