Add new model

Author Mitja Felicijan <mitja.felicijan@gmail.com> 2026-02-12 22:44:23 +0100
Committer Mitja Felicijan <mitja.felicijan@gmail.com> 2026-02-12 22:44:23 +0100
Commit 380a79e6cdf5cdc935eaf82a59c3358a8a2870d0 (patch)
-rw-r--r-- README.md 4
-rw-r--r-- context.txt 9
-rw-r--r-- models.h 24
-rw-r--r-- models.txt 4
4 files changed, 40 insertions, 1 deletions
diff --git a/README.md b/README.md
...
56 
```bash
 56 
```bash
57 
make clean
 57 
make clean
58 
```
 58 
```
 
  
 59 
  
 
  
 60 
## Reading material
 
  
 61 
  
 
  
 62 
- https://www.tinyllm.org/
diff --git a/context.txt b/context.txt 
  
 1 
Gandalf: wizard, Lord of the Rings, grey beard, staff, Istari, Grey Pilgrim, Mithrandir, fought Sauron, helped destroy One Ring.
 
  
 2 
  
 
  
 3 
Frodo: hobbit, Lord of the Rings, Bilbo's nephew, Shire, carried One Ring to Mount Doom, Fellowship of the Ring.
 
  
 4 
  
 
  
 5 
Example: Who is Gandalf? Gandalf is a wizard from The Lord of the Rings.
 
  
 6 
Example: Who is Frodo? Frodo is a hobbit from The Lord of the Rings.
 
  
 7 
Example: Who is Harry Potter? I don't have that information.
 
  
 8 
  
 
  
 9 
Answer this question. Use only the facts from above. If unknown, say "I don't have that information." Just give the answer, no prefix:
diff --git a/models.h b/models.h
...
31 
        .min_p = 0.05f,
 31 
        .min_p = 0.05f,
32 
        .seed = LLAMA_DEFAULT_SEED,
 32 
        .seed = LLAMA_DEFAULT_SEED,
33 
    },
 33 
    },
 
  
 34 
    {
 
  
 35 
        .name = "phi-4-mini-instruct",
 
  
 36 
        .filepath = "models/Phi-4-mini-instruct.Q2_K.gguf",
 
  
 37 
        .n_gpu_layers = 0,
 
  
 38 
        .use_mmap = false,
 
  
 39 
        .n_ctx = 131072,
 
  
 40 
        .n_batch = 4096,
 
  
 41 
        .embeddings = false,
 
  
 42 
        .temperature = 0.8f,
 
  
 43 
        .min_p = 0.05f,
 
  
 44 
        .seed = LLAMA_DEFAULT_SEED,
 
  
 45 
    },
 
  
 46 
    {
 
  
 47 
        .name = "tinyllama-1",
 
  
 48 
        .filepath = "models/TinyLlama-1.1B-intermediate-step-1431k-3T-Q2_K.gguf",
 
  
 49 
        .n_gpu_layers = 0,
 
  
 50 
        .use_mmap = false,
 
  
 51 
        .n_ctx = 2048,
 
  
 52 
        .n_batch = 4096,
 
  
 53 
        .embeddings = false,
 
  
 54 
        .temperature = 0.8f,
 
  
 55 
        .min_p = 0.05f,
 
  
 56 
        .seed = LLAMA_DEFAULT_SEED,
 
  
 57 
    },
34 
};
 58 
};
35 
  
 59 
  
36 
const model_config *get_model_by_name(const char *name) {
 60 
const model_config *get_model_by_name(const char *name) {
...
diff --git a/models.txt b/models.txt
1 
https://huggingface.co/Felladrin/gguf-flan-t5-small/resolve/main/flan-t5-small.F16.gguf?download=true
 1 
https://huggingface.co/Felladrin/gguf-flan-t5-small/resolve/main/flan-t5-small.F16.gguf
 
  
 2 
https://huggingface.co/MaziyarPanahi/Phi-4-mini-instruct-GGUF/resolve/main/Phi-4-mini-instruct.Q2_K.gguf
 
  
 3 
https://huggingface.co/andrijdavid/TinyLlama-1.1B-intermediate-step-1431k-3T-GGUF/resolve/main/TinyLlama-1.1B-intermediate-step-1431k-3T-Q2_K.gguf