Test g4 nano conversions/quantizations

Signed-off-by: Matt Rutkowski <mrutkows@us.ibm.com>

Build small-h ollama models

Signed-off-by: Matt Rutkowski <mrutkows@us.ibm.com>

Test g4 h builds

Signed-off-by: Matt Rutkowski <mrutkows@us.ibm.com>

Rebuild g4 instruct variants with new chat template

Signed-off-by: Matt Rutkowski <mrutkows@us.ibm.com>

config. for default Ollama quant. for small not updated

Signed-off-by: Matt Rutkowski <mrutkows@us.ibm.com>

Build only h-micro

Signed-off-by: Matt Rutkowski <mrutkows@us.ibm.com>

config. for default Ollama quant. for small not updated

Signed-off-by: Matt Rutkowski <mrutkows@us.ibm.com>

Try full g4 small build to ibm ollama knowing f16 will fail

Signed-off-by: Matt Rutkowski <mrutkows@us.ibm.com>

Re-run micro quants.

Signed-off-by: Matt Rutkowski <mrutkows@us.ibm.com>

Use production info

Signed-off-by: Matt Rutkowski <mrutkows@us.ibm.com>