Skip to content

Commit fb54d0f

Browse files
ankit-aglaweankit-aglawe
authored
gallery: add Parable Claude-Fable-5 agent-trace models (3B/4B/8B) (#10930)
1 parent 314a824 commit fb54d0f

1 file changed

Lines changed: 142 additions & 0 deletions

File tree

‎gallery/index.yaml‎

Lines changed: 142 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -1,4 +1,146 @@
11
---
2+
- name: "parable-granite-4.1-3b-claude-fable-5"
3+
url: "github:mudler/LocalAI/gallery/virtual.yaml@master"
4+
urls:
5+
- https://huggingface.co/AnkitAI/Parable-Granite-4.1-3B-Claude-Fable-5-GGUF
6+
description: |
7+
# Parable-Granite-4.1-3B-Claude-Fable-5
8+
9+
Granite 4.1 3B fine-tuned on genuine Claude Fable 5 and GPT-5.5 agent
10+
traces (planning, tool use, <think> reasoning from real agent sessions).
11+
Agent-flavored small model: terminal workflows, idiomatic code fixes,
12+
explanations. v2 recipe: completion-masked SFT, replay mix, seed-averaged
13+
weights. Published corpus and eval harness.
14+
license: "apache-2.0"
15+
tags:
16+
- llm
17+
- gguf
18+
- agent
19+
- coding
20+
- thinking
21+
- tool-use
22+
overrides:
23+
backend: llama-cpp
24+
function:
25+
automatic_tool_parsing_fallback: true
26+
grammar:
27+
disable: true
28+
known_usecases:
29+
- chat
30+
options:
31+
- use_jinja:true
32+
parameters:
33+
model: llama-cpp/models/Parable-Granite-4.1-3B-Claude-Fable-5-Q4_K_M/Parable-Granite-4.1-3B-Claude-Fable-5-GGUF-Q4_K_M.gguf
34+
template:
35+
use_tokenizer_template: true
36+
files:
37+
- filename: llama-cpp/models/Parable-Granite-4.1-3B-Claude-Fable-5-Q4_K_M/Parable-Granite-4.1-3B-Claude-Fable-5-GGUF-Q4_K_M.gguf
38+
uri: https://huggingface.co/AnkitAI/Parable-Granite-4.1-3B-Claude-Fable-5-GGUF/resolve/main/Parable-Granite-4.1-3B-Claude-Fable-5-GGUF-Q4_K_M.gguf
39+
sha256: 67dc7695d92939c713165761f115c9d892fdff74fcbd987c8bb453b9b8ab645d
40+
- name: "parable-qwen3-4b-claude-fable-5"
41+
url: "github:mudler/LocalAI/gallery/virtual.yaml@master"
42+
urls:
43+
- https://huggingface.co/AnkitAI/Parable-Qwen3-4B-Claude-Fable-5-GGUF
44+
description: |
45+
# Parable-Qwen3-4B-Claude-Fable-5
46+
47+
Qwen3 4B fine-tuned on genuine Claude Fable 5 agent traces. Thinking-mode
48+
reasoning, agent/terminal task flavor, tool-call formatting.
49+
license: "apache-2.0"
50+
tags:
51+
- llm
52+
- gguf
53+
- agent
54+
- coding
55+
- thinking
56+
overrides:
57+
backend: llama-cpp
58+
function:
59+
automatic_tool_parsing_fallback: true
60+
grammar:
61+
disable: true
62+
known_usecases:
63+
- chat
64+
options:
65+
- use_jinja:true
66+
parameters:
67+
model: llama-cpp/models/Parable-Qwen3-4B-Claude-Fable-5-Q4_K_M/Parable-Qwen3-4B-Claude-Fable-5-GGUF-Q4_K_M.gguf
68+
template:
69+
use_tokenizer_template: true
70+
files:
71+
- filename: llama-cpp/models/Parable-Qwen3-4B-Claude-Fable-5-Q4_K_M/Parable-Qwen3-4B-Claude-Fable-5-GGUF-Q4_K_M.gguf
72+
uri: https://huggingface.co/AnkitAI/Parable-Qwen3-4B-Claude-Fable-5-GGUF/resolve/main/Parable-Qwen3-4B-Claude-Fable-5-GGUF-Q4_K_M.gguf
73+
sha256: c6f991bd243fd1449d50a58a8de7d26bcad35d908bc0193e5a2c2c56ac6b8d5f
74+
- name: "parable-granite-4.1-8b-claude-fable-5"
75+
url: "github:mudler/LocalAI/gallery/virtual.yaml@master"
76+
urls:
77+
- https://huggingface.co/AnkitAI/Parable-Granite-4.1-8B-Claude-Fable-5-GGUF
78+
description: |
79+
# Parable-Granite-4.1-8B-Claude-Fable-5
80+
81+
Granite 4.1 8B fine-tuned on genuine Claude Fable 5 and GPT-5.5 agent
82+
traces. Strongest Parable model: multi-step scripts, configs, terminal
83+
workflows, <think> reasoning.
84+
license: "apache-2.0"
85+
tags:
86+
- llm
87+
- gguf
88+
- agent
89+
- coding
90+
- thinking
91+
- tool-use
92+
overrides:
93+
backend: llama-cpp
94+
function:
95+
automatic_tool_parsing_fallback: true
96+
grammar:
97+
disable: true
98+
known_usecases:
99+
- chat
100+
options:
101+
- use_jinja:true
102+
parameters:
103+
model: llama-cpp/models/Parable-Granite-4.1-8B-Claude-Fable-5-Q4_K_M/Parable-Granite-4.1-8B-Claude-Fable-5-GGUF-Q4_K_M.gguf
104+
template:
105+
use_tokenizer_template: true
106+
files:
107+
- filename: llama-cpp/models/Parable-Granite-4.1-8B-Claude-Fable-5-Q4_K_M/Parable-Granite-4.1-8B-Claude-Fable-5-GGUF-Q4_K_M.gguf
108+
uri: https://huggingface.co/AnkitAI/Parable-Granite-4.1-8B-Claude-Fable-5-GGUF/resolve/main/Parable-Granite-4.1-8B-Claude-Fable-5-GGUF-Q4_K_M.gguf
109+
sha256: 61a8133c344a0d0a00188395afe33c803e3b973cb4bbfd5ef1fa7110e80bc1c3
110+
- name: "parable-qwen3-8b-claude-fable-5"
111+
url: "github:mudler/LocalAI/gallery/virtual.yaml@master"
112+
urls:
113+
- https://huggingface.co/AnkitAI/Parable-Qwen3-8B-Claude-Fable-5-GGUF
114+
description: |
115+
# Parable-Qwen3-8B-Claude-Fable-5
116+
117+
Qwen3 8B fine-tuned on genuine Claude Fable 5 agent traces. Thinking-mode
118+
reasoning with agent/terminal flavor and tool-call formatting.
119+
license: "apache-2.0"
120+
tags:
121+
- llm
122+
- gguf
123+
- agent
124+
- coding
125+
- thinking
126+
overrides:
127+
backend: llama-cpp
128+
function:
129+
automatic_tool_parsing_fallback: true
130+
grammar:
131+
disable: true
132+
known_usecases:
133+
- chat
134+
options:
135+
- use_jinja:true
136+
parameters:
137+
model: llama-cpp/models/Parable-Qwen3-8B-Claude-Fable-5-Q4_K_M/Parable-Qwen3-8B-Claude-Fable-5-GGUF-Q4_K_M.gguf
138+
template:
139+
use_tokenizer_template: true
140+
files:
141+
- filename: llama-cpp/models/Parable-Qwen3-8B-Claude-Fable-5-Q4_K_M/Parable-Qwen3-8B-Claude-Fable-5-GGUF-Q4_K_M.gguf
142+
uri: https://huggingface.co/AnkitAI/Parable-Qwen3-8B-Claude-Fable-5-GGUF/resolve/main/Parable-Qwen3-8B-Claude-Fable-5-GGUF-Q4_K_M.gguf
143+
sha256: 956070afc8023b8665fe450842f7be76b505b53d142460fd9b588222f4e16112
2144
- &pocket-35b
3145
name: "pocket-35b"
4146
variants:

0 commit comments

Comments
 (0)