Merge remote-tracking branch 'origin/master'

.gitignore

@@ -2,4 +2,3 @@
 Dockerfile
 RUN_AIDER.md
 .ai-config
-.aider*

README.md

@@ -1,91 +1,16 @@
 # local_ai
 
-<p align="center">
-  <img src="logo.png" alt="local_ai: a laptop running Qwen, DeepSeek and Granite coding models" width="320">
-</p>
-
 Local coding AI for [aider](https://aider.chat): an Ollama container with a model
 chosen for *this* machine. Aider runs on the host and talks to the container.
 
 ```bash
 ./setup.sh            # analyse machine, pick model, generate Dockerfile + RUN_AIDER.md
 ./build-ai-image.sh   # build image (downloads the model)
-./run-ai-image.sh     # start container (port 11434, or the next free one)
+./run-ai-image.sh     # start container on 127.0.0.1:11434
 # then follow RUN_AIDER.md to start aider
 ```
 
 - `models.tsv` – model catalogue (edit to add models)
 - `Dockerfile.template` – container template; `setup.sh` fills in the model
 - `Dockerfile`, `.ai-config`, `RUN_AIDER.md` – generated per machine (git-ignored)
-- `./setup.sh --model <tag> --port <n> --gpu|--no-gpu --yes` for non-interactive use
-
-## What decides which model fits
-
-- **Memory is the limit.** A model must fit in VRAM (GPU) or RAM (CPU). `setup.sh` keeps
-  about 4 GB of RAM free for the OS, aider and your editor, and about 10% of VRAM.
-- **Fits in VRAM = fast.** The whole model on the GPU is by far the quickest option
-  (`fast (GPU)` in the menu).
-- **Bigger than VRAM = hybrid.** Ollama puts as many layers as fit on the GPU and runs the
-  rest on the CPU (`hybrid ~N% GPU`, a rough estimate). Faster than CPU-only, but the CPU
-  part sets the pace. Check the real split with `docker exec local-ai ollama ps`.
-- **CPU-only** works, at a few tokens/second for a 7B model. Mixture-of-Experts (MoE) models
-  such as `qwen3-coder:30b` only activate a few billion parameters per token, so they are
-  much faster on CPU than dense models of the same size, but still need the RAM for all of it.
-- **Context window** (4k to 32k tokens) is chosen from the memory budget and costs RAM/VRAM
-  too. Aider needs room for the repo map plus your files, so 8k is a practical minimum.
-- **Status in the menu:** `fits`, `tight` (fits with almost no headroom: close other apps),
-  `slow` (dense model without a GPU), `too big` (will swap or crash).
-- **Port:** the container uses 11434 like a normal Ollama. If that is taken (for example by a
-  host Ollama), `setup.sh` picks the next free port and writes it into `RUN_AIDER.md`.
-
-### Using an NVIDIA GPU
-
-Docker needs both the **NVIDIA driver** and **nvidia-container-toolkit**. `setup.sh` detects
-a GPU that is not usable yet and prints distro-specific install commands (also in
-`RUN_AIDER.md`), but only when a driver would actually help. Be careful with driver installs:
-
-- Take a snapshot first (Timeshift) and know how to reach the GRUB menu.
-- Prefer the distro's recommended driver (`ubuntu-drivers devices`), not an arbitrary version.
-- With Secure Boot on you must enrol the MOK key at the first reboot (`mokutil --sb-state`).
-- If the machine stops booting: boot an older kernel or add `nomodeset` in GRUB, then
-  remove the packages (`dpkg -l | awk '/^ii/ && /nvidia/ && !/linux-firmware-nvidia/ {print $2}' | xargs sudo dpkg --purge --force-depends`,
-  then `sudo update-initramfs -u`). No network is needed for that.
-- After a successful install, verify with `docker run --rm --gpus all ubuntu nvidia-smi` and
-  re-run `./setup.sh`.
-
-## Example configurations
-
-Speeds are rough orders of magnitude, not benchmarks. Your hardware will differ.
-
-### 1. CPU-only laptop, 16 GB RAM, no usable GPU
-
-`qwen2.5-coder:7b`, context 8k (this is what `setup.sh` picks on such a machine).
-
-| Pros | Cons |
-| --- | --- |
-| Works anywhere with Docker, no driver risk | A few tokens/second; long answers take a while |
-| Good quality for its size on everyday edits | First response is slow while the model loads into RAM |
-| Leaves ~9 GB RAM for the rest of the system | Keep prompts small and add only the files aider needs |
-
-### 2. Small GPU (about 4 GB VRAM, e.g. GTX 1650) with 16 GB RAM
-
-Either `qwen2.5-coder:3b` (fully on the GPU) or `qwen2.5-coder:7b` (hybrid, ~55% on the GPU).
-
-| Pros | Cons |
-| --- | --- |
-| 3B: fully in VRAM, fast and responsive | 3B: noticeably weaker; use `--edit-format whole` in aider |
-| 7B hybrid: better answers, still sped up by the GPU | 7B hybrid: speedup is modest, the CPU part dominates |
-| Both fit comfortably in 16 GB RAM | Needs the NVIDIA driver + container toolkit (driver install can break boot) |
-
-A reasonable approach: start with the 7B hybrid for quality and switch to the 3B for quick,
-simple edits. Re-run `./setup.sh` and `./build-ai-image.sh` to switch.
-
-### 3. Workstation with a 24 GB GPU and 32+ GB RAM
-
-`qwen2.5-coder:32b` (dense, best quality) or `qwen3-coder:30b` (MoE, faster).
-
-| Pros | Cons |
-| --- | --- |
-| Close to cloud-assistant quality for code | Large image (~20 GB) and long first build |
-| Fully in VRAM: fast, with room for a 16k+ context | The GPU is busy while the model is loaded |
-| MoE variant stays usable even if it spills into RAM | Needs the most expensive hardware; dense 32B is slow if it does not fit |
+- `./setup.sh --model <tag> --yes` for non-interactive use

models.tsv

@@ -3,15 +3,10 @@
 # cpu_ok: yes = usable speed without a GPU (small dense or MoE with few active params).
 # rank: higher = better quality for coding; the recommender picks the highest rank that fits.
 qwen2.5-coder:1.5b	1.0	2.0	yes	10	Tiny; very fast, weak edits. For very small machines.
-granite-code:3b	2.0	3.0	yes	15	IBM, Apache-2.0 (fully permissive); a bit weaker than Qwen at this size.
 qwen2.5-coder:3b	1.9	3.0	yes	20	Small and fast; OK for simple edits.
-granite-code:8b	4.6	6.2	yes	27	IBM, Apache-2.0 (fully permissive); solid mid-size coder.
 qwen2.5-coder:7b	4.7	6.5	yes	30	Best all-round choice for laptops / CPU-only.
-granite-code:20b	12	15	no	33	IBM, Apache-2.0 (fully permissive); dense, wants a GPU (~15 GB).
-deepseek-coder-v2:16b	8.9	11.5	yes	35	MoE (Mixture of Experts, 2.4B active): fast on CPU, needs ~12 GB RAM.
+deepseek-coder-v2:16b	8.9	11.5	yes	35	MoE (2.4B active): fast on CPU, needs ~12 GB RAM.
 qwen2.5-coder:14b	9.0	11.5	no	40	Noticeably smarter than 7B; wants a GPU (~12 GB VRAM).
-gpt-oss:20b	14	17	yes	50	MoE (Mixture of Experts) reasoning model; needs ~17 GB RAM or VRAM.
-granite-code:34b	19	23	no	55	IBM, Apache-2.0 (fully permissive); dense, needs a 24 GB GPU.
-devstral:24b	14	17	no	58	Mistral+All Hands, Apache-2.0; built for agentic multi-file edits, needs ~17 GB RAM or VRAM.
-qwen3-coder:30b	19	23	yes	60	MoE (Mixture of Experts, 3B active): strong coder, fast on CPU with 24+ GB RAM.
+gpt-oss:20b	14	17	yes	50	MoE reasoning model; needs ~17 GB RAM or VRAM.
+qwen3-coder:30b	19	23	yes	60	MoE (3B active): strong coder, fast on CPU with 24+ GB RAM.
 qwen2.5-coder:32b	20	24	no	70	Top-quality dense coder; needs a 24 GB GPU.

setup.sh

@@ -132,24 +132,15 @@ echo
 
 # ------------------------------------------------ memory budget / context
 # What a model must fit into: VRAM when a GPU is used, else RAM minus headroom
-# for the OS, aider and the editor. hard_budget/gpu_hard have almost no headroom -
-# a model only fitting there is "tight" (usable, but close other apps first).
+# for the OS, aider and the editor.
 budget=$(awk -v r="$ram_gb" 'BEGIN {b=r-4; if (b<1) b=1; printf "%.1f", b}')
-hard_budget=$(awk -v r="$ram_gb" 'BEGIN {b=r-1; if (b<1) b=1; printf "%.1f", b}')
 gpu_budget=$(awk -v v="$vram_gb" 'BEGIN {printf "%.1f", v*0.9}')
-gpu_hard=$(awk -v v="$vram_gb" 'BEGIN {printf "%.1f", v*0.98}')
 
 if   awk -v b="$budget" 'BEGIN{exit !(b<6)}';  then CONTEXT=4096
 elif awk -v b="$budget" 'BEGIN{exit !(b>=40)}'; then CONTEXT=32768
 elif awk -v b="$budget" 'BEGIN{exit !(b>=20)}'; then CONTEXT=16384
 else CONTEXT=8192; fi
 
-# Rough share (%) of a model's layers that fit in VRAM when a GPU is used and the
-# model is bigger than the VRAM budget (Ollama then offloads part of the layers).
-gpu_share() {
-  awk -v m="$1" -v b="$gpu_budget" -v u="$gpu_usable" 'BEGIN{ if(!u||m<=b){print 0; exit} s=int(b/m*100/5)*5; print (s<10?0:s) }'
-}
-
 # ---------------------------------------------------------------- catalogue
 mapfile -t rows < <(grep -v '^#' models.tsv | grep -v '^[[:space:]]*$' | sort -t$'\t' -k5,5n)
 tags=(); status=(); speed=(); rec=-1
@@ -157,16 +148,10 @@ for i in "${!rows[@]}"; do
   IFS=$'\t' read -r tag disk mem cpu_ok rank desc <<<"${rows[$i]}"
   tags[i]=$tag
   fits_ram=$(awk -v m="$mem" -v b="$budget" 'BEGIN{print (m<=b)}')
-  fits_hard_ram=$(awk -v m="$mem" -v b="$hard_budget" 'BEGIN{print (m<=b)}')
   fits_gpu=$(awk -v m="$mem" -v b="$gpu_budget" -v u="$gpu_usable" 'BEGIN{print (u && m<=b)}')
-  fits_hard_gpu=$(awk -v m="$mem" -v b="$gpu_hard" -v u="$gpu_usable" 'BEGIN{print (u && m<=b)}')
-  share=$(gpu_share "$mem")
-  if [ "$share" -gt 0 ]; then hyb="hybrid ~${share}% GPU"; else hyb=""; fi
   if [ "$fits_gpu" = 1 ]; then status[i]="fits"; speed[i]="fast (GPU)"
-  elif [ "$fits_hard_gpu" = 1 ]; then status[i]="tight"; speed[i]="tight (GPU)"
-  elif [ "$fits_ram" = 1 ] && [ "$cpu_ok" = yes ]; then status[i]="fits"; speed[i]="${hyb:-ok (CPU)}"
-  elif [ "$fits_ram" = 1 ]; then status[i]="slow"; speed[i]="slow ${hyb:-(CPU)}"
-  elif [ "$fits_hard_ram" = 1 ]; then status[i]="tight"; speed[i]="tight ${hyb:-(CPU)}"
+  elif [ "$fits_ram" = 1 ] && [ "$cpu_ok" = yes ]; then status[i]="fits"; speed[i]="ok (CPU)"
+  elif [ "$fits_ram" = 1 ]; then status[i]="slow"; speed[i]="slow (CPU)"
   else status[i]="too big"; speed[i]="-"; fi
   [ "${status[$i]}" = fits ] && rec=$i     # sorted by rank, so last "fits" wins
 done
@@ -212,25 +197,21 @@ fi
 # --------------------------------------------------------------------- menu
 if [ -z "$FORCE_MODEL" ]; then
   echo "=== Models (budget: ${budget} GB RAM$([ "$gpu_usable" = 1 ] && echo ", ${gpu_budget} GB VRAM")) ==="
-  printf "  %-3s %-24s %7s %7s  %-9s %-21s %s\n" "#" "model" "disk" "memory" "fit" "speed" ""
+  printf "  %-3s %-24s %7s %7s  %-9s %-11s %s\n" "#" "model" "disk" "memory" "fit" "speed" ""
   for i in "${!rows[@]}"; do
     IFS=$'\t' read -r tag disk mem cpu_ok rank desc <<<"${rows[$i]}"
     mark=" "; [ "$i" = "$rec" ] && mark="*"
-    printf " %s%-3s %-24s %5s GB %5s GB  %-9s %-21s %s\n" "$mark" "$((i+1))" "$tag" "$disk" "$mem" "${status[$i]}" "${speed[$i]}" "$desc"
+    printf " %s%-3s %-24s %5s GB %5s GB  %-9s %-11s %s\n" "$mark" "$((i+1))" "$tag" "$disk" "$mem" "${status[$i]}" "${speed[$i]}" "$desc"
   done
   echo
   echo "  * = recommended for this machine: ${tags[$rec]}"
-  [ "$gpu_usable" = 1 ] && echo "  hybrid ~N% GPU = model is bigger than VRAM; Ollama puts about N% of it on the GPU, the rest runs on the CPU (rough estimate)."
   echo
   if [ "$ASSUME_YES" = 1 ]; then choice=""; else read -r -p "Select model number [$((rec+1))]: " choice; fi
   choice=${choice:-$((rec+1))}
   [[ "$choice" =~ ^[0-9]+$ ]] && [ "$choice" -ge 1 ] && [ "$choice" -le "${#rows[@]}" ] || { echo "Invalid choice." >&2; exit 1; }
   sel=$((choice-1))
   MODEL=${tags[$sel]}
-  if [ "${status[$sel]}" = "tight" ]; then
-    echo "Warning: $MODEL is 'tight' on this machine (fits, but leaves little headroom - close other apps first)."
-    if [ "$ASSUME_YES" != 1 ]; then read -r -p "Continue anyway? [y/N] " a; [[ "$a" =~ ^[Yy] ]] || exit 1; fi
-  elif [ "${status[$sel]}" = "too big" ] || [ "${status[$sel]}" = "slow" ]; then
+  if [ "${status[$sel]}" = "too big" ] || [ "${status[$sel]}" = "slow" ]; then
     echo "Warning: $MODEL is '${status[$sel]}' on this machine (may swap, crawl, or crash)."
     if [ "$ASSUME_YES" != 1 ]; then read -r -p "Continue anyway? [y/N] " a; [[ "$a" =~ ^[Yy] ]] || exit 1; fi
   fi
@@ -274,10 +255,7 @@ case "$MODEL" in
   *:1.5b|*:3b) EDIT_HINT=$'\nSmall models struggle with diff-style edits; add `--edit-format whole`.\n' ;;
   *) EDIT_HINT="" ;;
 esac
-sel_mem=0; for r in "${rows[@]}"; do [ "$(cut -f1 <<<"$r")" = "$MODEL" ] && sel_mem=$(cut -f3 <<<"$r"); done
-sel_share=$(gpu_share "$sel_mem")
-if [ "$USE_GPU" = 1 ] && [ "$sel_share" -gt 0 ]; then SPEED="Hybrid mode: $MODEL is bigger than your VRAM, so roughly ${sel_share}% of it runs on the GPU and the rest on the CPU. Faster than CPU-only, slower than a model that fits entirely in VRAM. Check with: docker exec $CONTAINER_NAME ollama ps"
-elif [ "$USE_GPU" = 1 ]; then SPEED="The model fits in VRAM: expect fast responses."
+if [ "$USE_GPU" = 1 ]; then SPEED="Expect fast responses."
 else SPEED="CPU-only: expect a few tokens/second, and a slow first response while the model loads into RAM. Keep prompts and added files small."; fi
 
 {